Publications by Andreas Hotho

• [ 2023 ]
• [ 2022 ]
• [ 2021 ]
• [ 2020 ]
• [ 2019 ]
• [ 2018 ]
• [ 2017 ]
• [ 2016 ]
• [ 2015 ]
• [ 2014 ]
• [ 2013 ]
• [ 2012 ]
• [ 2011 ]
• [ 2010 ]
• [ 2009 ]
• [ 2008 ]
• [ 2007 ]
• [ 2006 ]
• [ 2005 ]
• [ 2004 ]
• [ 2003 ]
• [ 2002 ]
• [ 2001 ]
• [ 2000 ]

• 
  ConvMOS: climate model output statistics with deep learning M. Steininger; D. Abel; K. Ziegler; A. Krause; H. Paeth; A. Hotho in Data Mining and Knowledge Discovery (2023). 37(1) 136–166.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Climate models are the tool of choice for scientists researching climate change. Like all models they suffer from errors, particularly systematic and location-specific representation errors. One way to reduce these errors is model output statistics (MOS) where the model output is fitted to observational data with machine learning. In this work, we assess the use of convolutional Deep Learning climate MOS approaches and present the ConvMOS architecture which is specifically designed based on the observation that there are systematic and location-specific errors in the precipitation estimates of climate models. We apply ConvMOS models to the simulated precipitation of the regional climate model REMO, showing that a combination of per-location model parameters for reducing location-specific errors and global model parameters for reducing systematic errors is indeed beneficial for MOS performance. We find that ConvMOS models can reduce errors considerably and perform significantly better than three commonly used MOS approaches and plain ResNet and U-Net models in most cases. Our results show that non-linear MOS models underestimate the number of extreme precipitation events, which we alleviate by training models specialized towards extreme precipitation events with the imbalanced regression method DenseLoss. While we consider climate MOS, we argue that aspects of ConvMOS may also be beneficial in other domains with geospatial data, such as air pollution modeling or weather forecasts.

  @article{steininger2023convmos, abstract = {Climate models are the tool of choice for scientists researching climate change. Like all models they suffer from errors, particularly systematic and location-specific representation errors. One way to reduce these errors is model output statistics (MOS) where the model output is fitted to observational data with machine learning. In this work, we assess the use of convolutional Deep Learning climate MOS approaches and present the ConvMOS architecture which is specifically designed based on the observation that there are systematic and location-specific errors in the precipitation estimates of climate models. We apply ConvMOS models to the simulated precipitation of the regional climate model REMO, showing that a combination of per-location model parameters for reducing location-specific errors and global model parameters for reducing systematic errors is indeed beneficial for MOS performance. We find that ConvMOS models can reduce errors considerably and perform significantly better than three commonly used MOS approaches and plain ResNet and U-Net models in most cases. Our results show that non-linear MOS models underestimate the number of extreme precipitation events, which we alleviate by training models specialized towards extreme precipitation events with the imbalanced regression method DenseLoss. While we consider climate MOS, we argue that aspects of ConvMOS may also be beneficial in other domains with geospatial data, such as air pollution modeling or weather forecasts.}, author = {Steininger, Michael and Abel, Daniel and Ziegler, Katrin and Krause, Anna and Paeth, Heiko and Hotho, Andreas}, journal = {Data Mining and Knowledge Discovery}, keywords = {statistics}, number = 1, pages = {136–166}, title = {ConvMOS: climate model output statistics with deep learning}, volume = 37, year = 2023 }
• 
  TaylorPDENet: Learning PDEs from non-grid Data P. Heinisch; A. Dulny; A. Krause; A. Hotho (2023).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Modeling data obtained from dynamical systems has gained attention in recent years as a challenging task for machine learning models. Previous approaches assume the measurements to be distributed on a grid. However, for real-world applications like weather prediction, the observations are taken from arbitrary locations within the spatial domain. In this paper, we propose TaylorPDENet - a novel machine learning method that is designed to overcome this challenge. Our algorithm uses the multidimensional Taylor expansion of a dynamical system at each observation point to estimate the spatial derivatives to perform predictions. TaylorPDENet is able to accomplish two objectives simultaneously: accurately forecast the evolution of a complex dynamical system and explicitly reconstruct the underlying differential equation describing the system. We evaluate our model on a variety of advection-diffusion equations with different parameters and show that it performs similarly to equivalent approaches on grid-structured data while being able to process unstructured data as well.

  @misc{heinisch2023taylorpdenet, abstract = {Modeling data obtained from dynamical systems has gained attention in recent years as a challenging task for machine learning models. Previous approaches assume the measurements to be distributed on a grid. However, for real-world applications like weather prediction, the observations are taken from arbitrary locations within the spatial domain. In this paper, we propose TaylorPDENet - a novel machine learning method that is designed to overcome this challenge. Our algorithm uses the multidimensional Taylor expansion of a dynamical system at each observation point to estimate the spatial derivatives to perform predictions. TaylorPDENet is able to accomplish two objectives simultaneously: accurately forecast the evolution of a complex dynamical system and explicitly reconstruct the underlying differential equation describing the system. We evaluate our model on a variety of advection-diffusion equations with different parameters and show that it performs similarly to equivalent approaches on grid-structured data while being able to process unstructured data as well.}, author = {Heinisch, Paul and Dulny, Andrzej and Krause, Anna and Hotho, Andreas}, keywords = {neural-ode}, note = {cite arxiv:2306.14511}, title = {TaylorPDENet: Learning PDEs from non-grid Data}, year = 2023 }
• 
  Evaluating feature relevance XAI in network intrusion detection J. Tritscher; M. Wolf; A. Hotho; D. Schlör in The World Conference on eXplainable Artificial Intelligence (xAI 2023) - to appear (2023).
  • [ Abstract ]
  • [ BibTeX ]
  As machine learning models become increasingly complex, there is a growing need for explainability to understand and trust the decision-making processes. In the domain of network intrusion detection, post-hoc feature relevance explanations have been widely used to provide insight into the factors driving model decisions. However, recent research has highlighted challenges with these methods when applied to anomaly detection, which can vary in importance and impact depending on the application domain. In this paper, we investigate the challenges of post-hoc feature relevance explanations for network intrusion detection, a critical area for ensuring the security and integrity of computer networks. To gain a deeper understanding of these challenges for the application domain, we quantitatively and qualitatively investigate the popular feature relevance approach SHAP when explaining different network intrusion detection approaches. We conduct experiments to jointly evaluate detection quality and explainability, and explore the impact of replacement data, a commonly overlooked hyperparameter of post-hoc feature relevance approaches. We find that post-hoc XAI can provide high quality explanations, but requires a careful choice of its replacement data as default settings and common choices do not transfer across different detection models. Our study showcases the viability of post-hoc XAI for network intrusion detection systems, but highlights the need for rigorous evaluations of produced explanations.

  @article{tritscher2023evaluating, abstract = {As machine learning models become increasingly complex, there is a growing need for explainability to understand and trust the decision-making processes. In the domain of network intrusion detection, post-hoc feature relevance explanations have been widely used to provide insight into the factors driving model decisions. However, recent research has highlighted challenges with these methods when applied to anomaly detection, which can vary in importance and impact depending on the application domain. In this paper, we investigate the challenges of post-hoc feature relevance explanations for network intrusion detection, a critical area for ensuring the security and integrity of computer networks. To gain a deeper understanding of these challenges for the application domain, we quantitatively and qualitatively investigate the popular feature relevance approach SHAP when explaining different network intrusion detection approaches. We conduct experiments to jointly evaluate detection quality and explainability, and explore the impact of replacement data, a commonly overlooked hyperparameter of post-hoc feature relevance approaches. We find that post-hoc XAI can provide high quality explanations, but requires a careful choice of its replacement data as default settings and common choices do not transfer across different detection models. Our study showcases the viability of post-hoc XAI for network intrusion detection systems, but highlights the need for rigorous evaluations of produced explanations.}, author = {Tritscher, Julian and Wolf, Maximilian and Hotho, Andreas and Schlör, Daniel}, journal = {The World Conference on eXplainable Artificial Intelligence (xAI 2023) - to appear}, keywords = {xai}, title = {Evaluating feature relevance XAI in network intrusion detection}, year = 2023 }
• 
  Towards a Computational Analysis of Suspense: Detecting Dangerous Situations A. Zehe; J. Schröter; A. Hotho (2023).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Suspense is an important tool in storytelling to keep readers engaged and wanting to read more. However, it has so far not been studied extensively in Computational Literary Studies. In this paper, we focus on one of the elements authors can use to build up suspense: dangerous situations. We introduce a corpus of texts annotated with dangerous situations, distinguishing between 7 types of danger. Additionally, we annotate parts of the text that describe fear experienced by a character, regardless of the actual presence of danger. We present experiments towards the automatic detection of these situations, finding that unsupervised baseline methods can provide valuable signals for the detection, but more complex methods are necessary for further analysis. Not unexpectedly, the description of danger and fear often relies heavily on the context, both local (e.g., situations where danger is only mentioned, but not actually present) and global (e.g., "storm" being used in a literal sense in an adventure novel, but metaphorically in a romance novel).

  @misc{zehe2023towards, abstract = {Suspense is an important tool in storytelling to keep readers engaged and wanting to read more. However, it has so far not been studied extensively in Computational Literary Studies. In this paper, we focus on one of the elements authors can use to build up suspense: dangerous situations. We introduce a corpus of texts annotated with dangerous situations, distinguishing between 7 types of danger. Additionally, we annotate parts of the text that describe fear experienced by a character, regardless of the actual presence of danger. We present experiments towards the automatic detection of these situations, finding that unsupervised baseline methods can provide valuable signals for the detection, but more complex methods are necessary for further analysis. Not unexpectedly, the description of danger and fear often relies heavily on the context, both local (e.g., situations where danger is only mentioned, but not actually present) and global (e.g., "storm" being used in a literal sense in an adventure novel, but metaphorically in a romance novel).}, author = {Zehe, Albin and Schröter, Julian and Hotho, Andreas}, keywords = {word2vec}, note = {cite arxiv:2305.06818 Comment: 4 pages + references & appendix; work in progress}, title = {Towards a Computational Analysis of Suspense: Detecting Dangerous Situations}, year = 2023 }
• 
  Can Neural Networks Distinguish High-school Level Mathematical Concepts? S. Wankerl; A. Dulny; G. Goetz; A. Hotho (2023).
  • [ Abstract ]
  • [ BibTeX ]
  Processing symbolic mathematics algorithmically is an important field of research. It has applications in computer algebra systems and supports researchers as well as applied mathematicians in their daily work. Recently, exploring the ability of neural networks to grasp mathematical concepts has received special attention. One particularly complex task for neural networks is to understand the relation of two mathematical expressions to each other. Despite the advances in learning mathematical relationships, previous studies are limited by small-scale datasets, relatively simple formula construction by few axiomatic rules and even artifacts in the data. With this work, we aim at overcoming these limitations and provide a deeper insight into the representation power of neural networks for classifying mathematical relations. We introduce a novel data generation algorithm to allow for more complex formula compositions and fully include mathematical fields up to high-school level. We research several tree-based and sequential neural architectures for classifying mathematical relations and conduct a systematic analysis of the models against rule-based as well as neural baselines with a focus on varying data set complexity, generalization abilities, and understanding of syntactical patterns. Our findings highlight the effectiveness of tree-structured models for this task and show the potential of deep learning models to distinguish high-school level mathematical concepts.

  @conference{wankerl2023neural, abstract = {Processing symbolic mathematics algorithmically is an important field of research. It has applications in computer algebra systems and supports researchers as well as applied mathematicians in their daily work. Recently, exploring the ability of neural networks to grasp mathematical concepts has received special attention. One particularly complex task for neural networks is to understand the relation of two mathematical expressions to each other. Despite the advances in learning mathematical relationships, previous studies are limited by small-scale datasets, relatively simple formula construction by few axiomatic rules and even artifacts in the data. With this work, we aim at overcoming these limitations and provide a deeper insight into the representation power of neural networks for classifying mathematical relations. We introduce a novel data generation algorithm to allow for more complex formula compositions and fully include mathematical fields up to high-school level. We research several tree-based and sequential neural architectures for classifying mathematical relations and conduct a systematic analysis of the models against rule-based as well as neural baselines with a focus on varying data set complexity, generalization abilities, and understanding of syntactical patterns. Our findings highlight the effectiveness of tree-structured models for this task and show the potential of deep learning models to distinguish high-school level mathematical concepts.}, author = {Wankerl, Sebastian and Dulny, Andrzej and Goetz, Gerhard and Hotho, Andreas}, booktitle = {Accepted at ICDM}, keywords = {myown}, title = {Can Neural Networks Distinguish High-school Level Mathematical Concepts?}, year = 2023 }
• 
  Liquor-HGNN: A heterogeneous graph neural network for leakage detection in water distribution networks M. "Schaller; M. "Steininger; A. "Dulny; D. "Schlör; A. "Hotho in LWDA’23: Lernen, Wissen, Daten, Analysen. October 09--11, 2023, Marburg, Germany, (M. Leyer, Ed.) (2023).
  • [ BibTeX ]
  @article{schaller2023liquorhgnn, author = {"Schaller, Melanie" and "Steininger, Michael" and "Dulny, Andrzej" and "Schlör, Daniel" and "Hotho, Andreas"}, editor = {Leyer, M.}, journal = {LWDA'23: Lernen, Wissen, Daten, Analysen. October 09–11, 2023, Marburg, Germany}, keywords = {myown}, month = {Oktober}, title = {Liquor-HGNN: A heterogeneous graph neural network for leakage detection in water distribution networks}, year = 2023 }
• 
  CapsKG: Enabling Continual Knowledge Integration in Language Models for Automatic Knowledge Graph Completion J. Omeliyanenko; A. Zehe; A. Hotho; D. Schlör in International Semantic Web Conference ISWC 2023, to appear (2023).
  • [ Abstract ]
  • [ BibTeX ]
  Automated completion of knowledge graphs is a popular topic in the Semantic Web community that aims to automatically and continuously integrate new appearing knowledge into knowledge graphs using artificial intelligence. Recently, approaches that leverage implicit knowledge from language models for this task have shown promising re- sults. However, by fine-tuning language models directly to the domain of knowledge graphs, models forget their original language representation and associated knowledge. An existing solution to address this issue is a trainable adapter, which is integrated into a frozen language model to extract the relevant knowledge without altering the model itself. How- ever, this constrains the generalizability to the specific extraction task and by design requires new and independent adapters to be trained for new knowledge extraction tasks. This effectively prevents the model from benefiting from existing knowledge incorporated in previously trained adapters. In this paper, we propose to combine the benefits of adapters for knowl- edge graph completion with the idea of integrating capsules, introduced in the field of continual learning. This allows the continuous integra- tion of knowledge into a joint model by sharing and reusing previously trained capsules. We find that our approach outperforms solutions using traditional adapters, while requiring notably fewer parameters for con- tinuous knowledge integration. Moreover, we show that this architecture benefits significantly from knowledge sharing in low-resource situations, outperforming adapter-based models on the task of link prediction.

  @article{omeliyanenko2023capskg, abstract = {Automated completion of knowledge graphs is a popular topic in the Semantic Web community that aims to automatically and continuously integrate new appearing knowledge into knowledge graphs using artificial intelligence. Recently, approaches that leverage implicit knowledge from language models for this task have shown promising re- sults. However, by fine-tuning language models directly to the domain of knowledge graphs, models forget their original language representation and associated knowledge. An existing solution to address this issue is a trainable adapter, which is integrated into a frozen language model to extract the relevant knowledge without altering the model itself. How- ever, this constrains the generalizability to the specific extraction task and by design requires new and independent adapters to be trained for new knowledge extraction tasks. This effectively prevents the model from benefiting from existing knowledge incorporated in previously trained adapters. In this paper, we propose to combine the benefits of adapters for knowl- edge graph completion with the idea of integrating capsules, introduced in the field of continual learning. This allows the continuous integra- tion of knowledge into a joint model by sharing and reusing previously trained capsules. We find that our approach outperforms solutions using traditional adapters, while requiring notably fewer parameters for con- tinuous knowledge integration. Moreover, we show that this architecture benefits significantly from knowledge sharing in low-resource situations, outperforming adapter-based models on the task of link prediction.}, author = {Omeliyanenko, Janna and Zehe, Albin and Hotho, Andreas and Schlör, Daniel}, journal = {International Semantic Web Conference ISWC 2023, to appear}, keywords = {myown}, title = {CapsKG: Enabling Continual Knowledge Integration in Language Models for Automatic Knowledge Graph Completion}, year = 2023 }
• 
  Swarming Detection in Smart Beehives Using Auto Encoders for Audio Data P. Janetzky; M. Schaller; A. Krause; A. Hotho (2023). 1–5.
  • [ Abstract ]
  • [ BibTeX ]
  • [ DOI ]
  —Swarming is the natural mechanism by which bee colonies reproduce, but for beekeepers it is a challenge. Precision beekeeping can aid their work through early notifications about impending swarms. In this work, we focus on identifying swarms and their early indicators in audio data captured from a smart beehive. The challenge with such domain-specific data is the low availability of labelled samples, the strong label imbalance, and the recording of undesired sources. We approach this challenge through a two-step setup: First, we use an auto encoder network to detect sounds from mechanical sources and then use it to clean data. Secondly, on the cleaned data we then employ a second network to identify event-related bee sounds. Using spectrogram features, our networks are able to reach a balanced accuracy score of more than 99 % in the detection of special bee events. The findings of this initial study can serve as the starting point for further research on handling imbalanced data collections from smart, remote sensor environments that also contain undesired signals.

  @inproceedings{janetzky2023swarming, abstract = {—Swarming is the natural mechanism by which bee colonies reproduce, but for beekeepers it is a challenge. Precision beekeeping can aid their work through early notifications about impending swarms. In this work, we focus on identifying swarms and their early indicators in audio data captured from a smart beehive. The challenge with such domain-specific data is the low availability of labelled samples, the strong label imbalance, and the recording of undesired sources. We approach this challenge through a two-step setup: First, we use an auto encoder network to detect sounds from mechanical sources and then use it to clean data. Secondly, on the cleaned data we then employ a second network to identify event-related bee sounds. Using spectrogram features, our networks are able to reach a balanced accuracy score of more than 99 % in the detection of special bee events. The findings of this initial study can serve as the starting point for further research on handling imbalanced data collections from smart, remote sensor environments that also contain undesired signals.}, author = {Janetzky, Pascal and Schaller, Melanie and Krause, Anna and Hotho, Andreas}, booktitle = {2023 30th International Conference on Systems, Signals and Image Processing (IWSSIP)}, keywords = {we4bee_audio}, pages = {1-5}, title = {Swarming Detection in Smart Beehives Using Auto Encoders for Audio Data}, year = 2023 }
• 
  Automatic Speech Detection on a Smart Beehive’s Raspberry Pi P. Janetzky; P. Lissmann; A. Hotho; A. Krause (2023).
  • [ BibTeX ]
  @inproceedings{janetzky2023automatic, author = {Janetzky, Pascal and Lissmann, Philipp and Hotho, Andreas and Krause, Anna}, booktitle = {LWDA 2023 - Learning, Knowledge, Data, and Analysis (accepted)}, keywords = {speechdetection}, title = {Automatic Speech Detection on a Smart Beehive's Raspberry Pi}, year = 2023 }
• 
  Occupational Fraud Detection through Agent-based Data Generation J. Tritscher; A. Roos; D. Schlör; A. Hotho; A. Krause in The 8th Workshop on MIning DAta for financial applicationS MIDAS 2023 - to appear (2023).
  • [ Abstract ]
  • [ BibTeX ]
  Occupational fraud is an increasing concern for enterprises that is estimated to cause losses of around 5% of company revenue each year. With the increasing data tracked by companies through enterprise resource planning systems, recent research has taken interest in the automated detection of occupational fraud. Automated detection is however hindered by the unavailability of labeled fraud cases which require known occupational frauds within company data and costly expert annotation. Even despite the existence of anomaly detection methods that can be trained on unsupervised data, selecting the ideal preprocessing techniques, the most suitable model, and the optimal hyperparameters necessitates the availability of labeled data for evaluation purposes. To alleviate this issue, we propose to use simulation through multi-agent systems for generating business processes according to best practices from economics and creating labeled synthetic data that closely matches a given unlabeled real-world dataset. We extend an existing simulation by incorporating functionality for including, tracking and automatic labeling of occupational fraud cases. Using this simulation, we propose a framework that decides on important design choices for fraud detection models in enterprise resource planning data and does not require labeled real-world data. We demonstrate in multiple experiments that the framework can aid automated occupational fraud detection through data generation.

  @article{tritscher2023occupational, abstract = {Occupational fraud is an increasing concern for enterprises that is estimated to cause losses of around 5% of company revenue each year. With the increasing data tracked by companies through enterprise resource planning systems, recent research has taken interest in the automated detection of occupational fraud. Automated detection is however hindered by the unavailability of labeled fraud cases which require known occupational frauds within company data and costly expert annotation. Even despite the existence of anomaly detection methods that can be trained on unsupervised data, selecting the ideal preprocessing techniques, the most suitable model, and the optimal hyperparameters necessitates the availability of labeled data for evaluation purposes. To alleviate this issue, we propose to use simulation through multi-agent systems for generating business processes according to best practices from economics and creating labeled synthetic data that closely matches a given unlabeled real-world dataset. We extend an existing simulation by incorporating functionality for including, tracking and automatic labeling of occupational fraud cases. Using this simulation, we propose a framework that decides on important design choices for fraud detection models in enterprise resource planning data and does not require labeled real-world data. We demonstrate in multiple experiments that the framework can aid automated occupational fraud detection through data generation.}, author = {Tritscher, Julian and Roos, Alexander and Schlör, Daniel and Hotho, Andreas and Krause, Anna}, journal = {The 8th Workshop on MIning DAta for financial applicationS MIDAS 2023 - to appear}, keywords = {simulation}, title = {Occupational Fraud Detection through Agent-based Data Generation}, year = 2023 }
• 
  Higher-Order DeepTrails: Unified Approach to *Trails T. Koopmann; J. Pfister; A. Markus; A. Carolus; C. Wienrich; A. Hotho in {CEUR} Workshop Proceedings, M. Leyer, J. Wichmann (Eds.) (2023).
  • [ BibTeX ]
  @inproceedings{koopmann2023higherorder, author = {Koopmann, Tobias and Pfister, Jan and Markus, André and Carolus, Astrid and Wienrich, Carolin and Hotho, Andreas}, booktitle = {Proceedings of the {LWDA} 2023 Workshops: FGWM, FGKD, and FGDB, Marburg (Germany), Oktober 9-11th, 2023}, editor = {Leyer, Michael and Wichmann, Johannes}, keywords = {myown}, publisher = {CEUR-WS.org}, series = {{CEUR} Workshop Proceedings}, title = {Higher-Order DeepTrails: Unified Approach to *Trails}, year = 2023 }
• 
  DynaBench: A Benchmark Dataset for Learning Dynamical Systems from Low-Resolution Data A. Dulny; A. Hotho; A. Krause D. Koutra, C. Plant, M. Gomez Rodriguez, E. Baralis, F. Bonchi (Eds.) (2023). 438–455.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Previous work on learning physical systems from data has focused on high-resolution grid-structured measurements. However, real-world knowledge of such systems (e.g. weather data) relies on sparsely scattered measuring stations. In this paper, we introduce a novel simulated benchmark dataset, DynaBench, for learning dynamical systems directly from sparsely scattered data without prior knowledge of the equations. The dataset focuses on predicting the evolution of a dynamical system from low-resolution, unstructured measurements. We simulate six different partial differential equations covering a variety of physical systems commonly used in the literature and evaluate several machine learning models, including traditional graph neural networks and point cloud processing models, with the task of predicting the evolution of the system. The proposed benchmark dataset is expected to advance the state of art as an out-of-the-box easy-to-use tool for evaluating models in a setting where only unstructured low-resolution observations are available. The benchmark is available at https://professor-x.de/dynabench.

  @inproceedings{dulny2023dynabench, abstract = {Previous work on learning physical systems from data has focused on high-resolution grid-structured measurements. However, real-world knowledge of such systems (e.g. weather data) relies on sparsely scattered measuring stations. In this paper, we introduce a novel simulated benchmark dataset, DynaBench, for learning dynamical systems directly from sparsely scattered data without prior knowledge of the equations. The dataset focuses on predicting the evolution of a dynamical system from low-resolution, unstructured measurements. We simulate six different partial differential equations covering a variety of physical systems commonly used in the literature and evaluate several machine learning models, including traditional graph neural networks and point cloud processing models, with the task of predicting the evolution of the system. The proposed benchmark dataset is expected to advance the state of art as an out-of-the-box easy-to-use tool for evaluating models in a setting where only unstructured low-resolution observations are available. The benchmark is available at https://professor-x.de/dynabench.}, address = {Cham}, author = {Dulny, Andrzej and Hotho, Andreas and Krause, Anna}, booktitle = {Machine Learning and Knowledge Discovery in Databases: Research Track}, editor = {Koutra, Danai and Plant, Claudia and Gomez Rodriguez, Manuel and Baralis, Elena and Bonchi, Francesco}, keywords = {neural-ode}, pages = {438–455}, publisher = {Springer Nature Switzerland}, title = {DynaBench: A Benchmark Dataset for Learning Dynamical Systems from Low-Resolution Data}, year = 2023 }
• 
  Long-Term Effects of Perceived Friendship with Intelligent Voice Assistants on Usage Behavior, User Experience, and Social Perceptions C. Wienrich; A. Carolus; A. Markus; Y. Augustin; J. Pfister; A. Hotho in Computers (2023). 12(4)
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Social patterns and roles can develop when users talk to intelligent voice assistants (IVAs) daily. The current study investigates whether users assign different roles to devices and how this affects their usage behavior, user experience, and social perceptions. Since social roles take time to establish, we equipped 106 participants with Alexa or Google assistants and some smart home devices and observed their interactions for nine months. We analyzed diverse subjective (questionnaire) and objective data (interaction data). By combining social science and data science analyses, we identified two distinct clusters—users who assigned a friendship role to IVAs over time and users who did not. Interestingly, these clusters exhibited significant differences in their usage behavior, user experience, and social perceptions of the devices. For example, participants who assigned a role to IVAs attributed more friendship to them used them more frequently, reported more enjoyment during interactions, and perceived more empathy for IVAs. In addition, these users had distinct personal requirements, for example, they reported more loneliness. This study provides valuable insights into the role-specific effects and consequences of voice assistants. Recent developments in conversational language models such as ChatGPT suggest that the findings of this study could make an important contribution to the design of dialogic human–AI interactions.

  @article{wienrich2023longterm, abstract = {Social patterns and roles can develop when users talk to intelligent voice assistants (IVAs) daily. The current study investigates whether users assign different roles to devices and how this affects their usage behavior, user experience, and social perceptions. Since social roles take time to establish, we equipped 106 participants with Alexa or Google assistants and some smart home devices and observed their interactions for nine months. We analyzed diverse subjective (questionnaire) and objective data (interaction data). By combining social science and data science analyses, we identified two distinct clusters—users who assigned a friendship role to IVAs over time and users who did not. Interestingly, these clusters exhibited significant differences in their usage behavior, user experience, and social perceptions of the devices. For example, participants who assigned a role to IVAs attributed more friendship to them used them more frequently, reported more enjoyment during interactions, and perceived more empathy for IVAs. In addition, these users had distinct personal requirements, for example, they reported more loneliness. This study provides valuable insights into the role-specific effects and consequences of voice assistants. Recent developments in conversational language models such as ChatGPT suggest that the findings of this study could make an important contribution to the design of dialogic human–AI interactions.}, author = {Wienrich, Carolin and Carolus, Astrid and Markus, André and Augustin, Yannik and Pfister, Jan and Hotho, Andreas}, journal = {Computers}, keywords = {myown}, number = 4, title = {Long-Term Effects of Perceived Friendship with Intelligent Voice Assistants on Usage Behavior, User Experience, and Social Perceptions}, volume = 12, year = 2023 }
• 
  Feature relevance XAI in anomaly detection: Reviewing approaches and challenges J. Tritscher; A. Krause; A. Hotho in Frontiers in Artificial Intelligence (2023). 6
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  With complexity of artificial intelligence systems increasing continuously in past years, studies to explain these complex systems have grown in popularity. While much work has focused on explaining artificial intelligence systems in popular domains such as classification and regression, explanations in the area of anomaly detection have only recently received increasing attention from researchers. In particular, explaining singular model decisions of a complex anomaly detector by highlighting which inputs were responsible for a decision, commonly referred to as local post-hoc feature relevance, has lately been studied by several authors. In this paper, we systematically structure these works based on their access to training data and the anomaly detection model, and provide a detailed overview of their operation in the anomaly detection domain. We demonstrate their performance and highlight their limitations in multiple experimental showcases, discussing current challenges and opportunities for future work in feature relevance XAI for anomaly detection.

  @article{tritscher2023feature, abstract = {With complexity of artificial intelligence systems increasing continuously in past years, studies to explain these complex systems have grown in popularity. While much work has focused on explaining artificial intelligence systems in popular domains such as classification and regression, explanations in the area of anomaly detection have only recently received increasing attention from researchers. In particular, explaining singular model decisions of a complex anomaly detector by highlighting which inputs were responsible for a decision, commonly referred to as local post-hoc feature relevance, has lately been studied by several authors. In this paper, we systematically structure these works based on their access to training data and the anomaly detection model, and provide a detailed overview of their operation in the anomaly detection domain. We demonstrate their performance and highlight their limitations in multiple experimental showcases, discussing current challenges and opportunities for future work in feature relevance XAI for anomaly detection.}, author = {Tritscher, Julian and Krause, Anna and Hotho, Andreas}, journal = {Frontiers in Artificial Intelligence}, keywords = {xai}, month = {feb}, title = {Feature relevance XAI in anomaly detection: Reviewing approaches and challenges}, volume = 6, year = 2023 }
• 
  Zero-Shot Clickbait Spoiling by Rephrasing Titles as Questions D. Wangsadirdja; J. Pfister; K. Kobs; A. Hotho (2023). 1090–1095.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  In this paper, we describe our approach to the clickbait spoiling task of SemEval 2023. The core idea behind our system is to leverage pre-trained models capable of Question Answering (QA) to extract the spoiler from article texts based on the clickbait title without any task-specific training. Since oftentimes, these titles are not phrased as questions, we automatically rephrase the clickbait titles as questions in order to better suit the pretraining task of the QA-capable models. Also, to fit as much relevant context into the model's limited input size as possible, we propose to reorder the sentences by their relevance using a semantic similarity model. Finally, we evaluate QA as well as text generation models (via prompting) to extract the spoiler from the text.Based on the validation data, our final model selects each of these components depending on the spoiler type and achieves satisfactory zero-shot results. The ideas described in this paper can easily be applied in fine-tuning settings.

  @inproceedings{wangsadirdja2023zeroshot, abstract = {In this paper, we describe our approach to the clickbait spoiling task of SemEval 2023. The core idea behind our system is to leverage pre-trained models capable of Question Answering (QA) to extract the spoiler from article texts based on the clickbait title without any task-specific training. Since oftentimes, these titles are not phrased as questions, we automatically rephrase the clickbait titles as questions in order to better suit the pretraining task of the QA-capable models. Also, to fit as much relevant context into the model's limited input size as possible, we propose to reorder the sentences by their relevance using a semantic similarity model. Finally, we evaluate QA as well as text generation models (via prompting) to extract the spoiler from the text.Based on the validation data, our final model selects each of these components depending on the spoiler type and achieves satisfactory zero-shot results. The ideas described in this paper can easily be applied in fine-tuning settings.}, address = {Toronto, Canada}, author = {Wangsadirdja, Dirk and Pfister, Jan and Kobs, Konstantin and Hotho, Andreas}, booktitle = {Proceedings of the The 17th International Workshop on Semantic Evaluation (SemEval-2023)}, keywords = {zeroshot}, month = {jul}, pages = {1090–1095}, publisher = {Association for Computational Linguistics}, title = {Zero-Shot Clickbait Spoiling by Rephrasing Titles as Questions}, year = 2023 }

• 
  On Background Bias in Deep Metric Learning K. Kobs; A. Hotho in CoRR (2022). abs/2210.01615
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2210-01615, author = {Kobs, Konstantin and Hotho, Andreas}, journal = {CoRR}, keywords = {myown}, title = {On Background Bias in Deep Metric Learning}, volume = {abs/2210.01615}, year = 2022 }
• 
  Semi-unsupervised Learning for Time Series Classification P. Davidson; M. Steininger; A. Huhn; A. Krause; A. Hotho in CoRR (2022). abs/2207.03119
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2207-03119, author = {Davidson, Padraig and Steininger, Michael and Huhn, Andr{\'{e}} and Krause, Anna and Hotho, Andreas}, journal = {CoRR}, keywords = {myown}, title = {Semi-unsupervised Learning for Time Series Classification}, volume = {abs/2207.03119}, year = 2022 }
• 
  The Impact of Different System Call Representations on Intrusion Detection S. Wunderlich; M. Ring; D. Landes; A. Hotho in Log. J. {IGPL} (2022). 30(2) 239–251.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/igpl/WunderlichRLH22, author = {Wunderlich, Sarah and Ring, Markus and Landes, Dieter and Hotho, Andreas}, journal = {Log. J. {IGPL}}, keywords = {myown}, number = 2, pages = {239–251}, title = {The Impact of Different System Call Representations on Intrusion Detection}, volume = 30, year = 2022 }
• 
  Anomaly Detection in Beehives: An Algorithm Comparison P. Davidson; M. Steininger; F. Lautenschlager; A. Krause; A. Hotho A. Ahrens, R. V. Prasad, C. Benavente-Peces, N. Ansari (Eds.) (2022). 1–20.
  • [ Abstract ]
  • [ BibTeX ]
  Sensor-equipped beehives allow monitoring the living conditions of bees. Machine learning models can use the data of such hives to learn behavioral patterns and find anomalous events. One type of event that is of particular interest to apiarists for economical reasons is bee swarming. Other events of interest are behavioral anomalies from illness and technical anomalies, e.g. sensor failure. Beekeepers can be supported by suitable machine learning models which can detect these events.

  @inproceedings{10.1007/978-3-031-17718-7_1, abstract = {Sensor-equipped beehives allow monitoring the living conditions of bees. Machine learning models can use the data of such hives to learn behavioral patterns and find anomalous events. One type of event that is of particular interest to apiarists for economical reasons is bee swarming. Other events of interest are behavioral anomalies from illness and technical anomalies, e.g. sensor failure. Beekeepers can be supported by suitable machine learning models which can detect these events.}, address = {Cham}, author = {Davidson, Padraig and Steininger, Michael and Lautenschlager, Florian and Krause, Anna and Hotho, Andreas}, booktitle = {Sensor Networks}, editor = {Ahrens, Andreas and Prasad, RangaRao Venkatesha and Benavente-Peces, C{\'e}sar and Ansari, Nirwan}, keywords = {myown}, pages = {1–20}, publisher = {Springer International Publishing}, title = {Anomaly Detection in Beehives: An Algorithm Comparison}, year = 2022 }
• 
  One Graph to Rule them All: Using {NLP} and Graph Neural Networks to analyse Tolkien’s Legendarium V. Perri; L. Qarkaxhija; A. Zehe; A. Hotho; I. Scholtes in {CEUR} Workshop Proceedings, F. Karsdorp, K. L. Nielbo (Eds.) (2022). (Vol. 3290) 291–317.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{DBLP:conf/chr/PerriQZHS22, author = {Perri, Vincenzo and Qarkaxhija, Lisi and Zehe, Albin and Hotho, Andreas and Scholtes, Ingo}, booktitle = {Proceedings of the Computational Humanities Research Conference 2022, {CHR} 2022, Antwerp, Belgium, December 12-14, 2022}, editor = {Karsdorp, Folgert and Nielbo, Kristoffer L.}, keywords = {myown}, pages = {291–317}, publisher = {CEUR-WS.org}, series = {{CEUR} Workshop Proceedings}, title = {One Graph to Rule them All: Using {NLP} and Graph Neural Networks to analyse Tolkien's Legendarium}, volume = 3290, year = 2022 }
• 
  Sequential Item Recommendation in the {MOBA} Game Dota 2 A. Dallmann; J. Kohlmann; D. Zoller; A. Hotho in CoRR (2022). abs/2201.08724
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2201-08724, author = {Dallmann, Alexander and Kohlmann, Johannes and Zoller, Daniel and Hotho, Andreas}, journal = {CoRR}, keywords = {myown}, title = {Sequential Item Recommendation in the {MOBA} Game Dota 2}, volume = {abs/2201.08724}, year = 2022 }
• 
  CLIP knows image aesthetics S. Hentschel; K. Kobs; A. Hotho in Frontiers in artificial intelligence (2022). 5 976235.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Most Image Aesthetic Assessment (IAA) methods use a pretrained ImageNet classification model as a base to fine-tune. We hypothesize that content classification is not an optimal pretraining task for IAA, since the task discourages the extraction of features that are useful for IAA, e.g., composition, lighting, or style. On the other hand, we argue that the Contrastive Language-Image Pretraining (CLIP) model is a better base for IAA models, since it has been trained using natural language supervision. Due to the rich nature of language, CLIP needs to learn a broad range of image features that correlate with sentences describing the image content, composition, environments, and even subjective feelings about the image. While it has been shown that CLIP extracts features useful for content classification tasks, its suitability for tasks that require the extraction of style-based features like IAA has not yet been shown. We test our hypothesis by conducting a three-step study, investigating the usefulness of features extracted by CLIP compared to features obtained from the last layer of a comparable ImageNet classification model. In each step, we get more computationally expensive. First, we engineer natural language prompts that let CLIP assess an image's aesthetic without adjusting any weights in the model. To overcome the challenge that CLIP's prompting only is applicable to classification tasks, we propose a simple but effective strategy to convert multiple prompts to a continuous scalar as required when predicting an image's mean aesthetic score. Second, we train a linear regression on the AVA dataset using image features obtained by CLIP's image encoder. The resulting model outperforms a linear regression trained on features from an ImageNet classification model. It also shows competitive performance with fully fine-tuned networks based on ImageNet, while only training a single layer. Finally, by fine-tuning CLIP's image encoder on the AVA dataset, we show that CLIP only needs a fraction of training epochs to converge, while also performing better than a fine-tuned ImageNet model. Overall, our experiments suggest that CLIP is better suited as a base model for IAA methods than ImageNet pretrained networks.

  @article{PMID:36504688, abstract = {Most Image Aesthetic Assessment (IAA) methods use a pretrained ImageNet classification model as a base to fine-tune. We hypothesize that content classification is not an optimal pretraining task for IAA, since the task discourages the extraction of features that are useful for IAA, e.g., composition, lighting, or style. On the other hand, we argue that the Contrastive Language-Image Pretraining (CLIP) model is a better base for IAA models, since it has been trained using natural language supervision. Due to the rich nature of language, CLIP needs to learn a broad range of image features that correlate with sentences describing the image content, composition, environments, and even subjective feelings about the image. While it has been shown that CLIP extracts features useful for content classification tasks, its suitability for tasks that require the extraction of style-based features like IAA has not yet been shown. We test our hypothesis by conducting a three-step study, investigating the usefulness of features extracted by CLIP compared to features obtained from the last layer of a comparable ImageNet classification model. In each step, we get more computationally expensive. First, we engineer natural language prompts that let CLIP assess an image's aesthetic without adjusting any weights in the model. To overcome the challenge that CLIP's prompting only is applicable to classification tasks, we propose a simple but effective strategy to convert multiple prompts to a continuous scalar as required when predicting an image's mean aesthetic score. Second, we train a linear regression on the AVA dataset using image features obtained by CLIP's image encoder. The resulting model outperforms a linear regression trained on features from an ImageNet classification model. It also shows competitive performance with fully fine-tuned networks based on ImageNet, while only training a single layer. Finally, by fine-tuning CLIP's image encoder on the AVA dataset, we show that CLIP only needs a fraction of training epochs to converge, while also performing better than a fine-tuned ImageNet model. Overall, our experiments suggest that CLIP is better suited as a base model for IAA methods than ImageNet pretrained networks.}, author = {Hentschel, Simon and Kobs, Konstantin and Hotho, Andreas}, journal = {Frontiers in artificial intelligence}, keywords = {myown}, pages = 976235, title = {CLIP knows image aesthetics}, volume = 5, year = 2022 }
• 
  Towards Responsible Medical Diagnostics Recommendation Systems D. Schl{\"{o}}r; A. Hotho in CoRR (2022). abs/2209.03760
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2209-03760, author = {Schl{\"{o}}r, Daniel and Hotho, Andreas}, journal = {CoRR}, keywords = {myown}, title = {Towards Responsible Medical Diagnostics Recommendation Systems}, volume = {abs/2209.03760}, year = 2022 }
• 
  Comparison of Data Representations and Machine Learning Architectures for User Identification on Arbitrary Motion Sequences C. Schell; A. Hotho; M. E. Latoschik in CoRR (2022). abs/2210.00527
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2210-00527, author = {Schell, Christian and Hotho, Andreas and Latoschik, Marc Erich}, journal = {CoRR}, keywords = {myown}, title = {Comparison of Data Representations and Machine Learning Architectures for User Identification on Arbitrary Motion Sequences}, volume = {abs/2210.00527}, year = 2022 }
• 
  Point me to your Opinion, {S}en{P}oi J. Pfister; S. Wankerl; A. Hotho (2022). 1313–1323.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Structured Sentiment Analysis is the task of extracting sentiment tuples in a graph structure commonly from review texts. We adapt the Aspect-Based Sentiment Analysis pointer network BARTABSA to model this tuple extraction as a sequence prediction task and extend their output grammar to account for the increased complexity of Structured Sentiment Analysis. To predict structured sentiment tuples in languages other than English we swap BART for a multilingual mT5 and introduce a novel Output Length Regularization to mitigate overfitting to common target sequence lengths, thereby improving the performance of the model by up to 70{\%}. We evaluate our approach on seven datasets in five languages including a zero shot crosslingual setting.

  @inproceedings{pfister-etal-2022-senpoi, abstract = {Structured Sentiment Analysis is the task of extracting sentiment tuples in a graph structure commonly from review texts. We adapt the Aspect-Based Sentiment Analysis pointer network BARTABSA to model this tuple extraction as a sequence prediction task and extend their output grammar to account for the increased complexity of Structured Sentiment Analysis. To predict structured sentiment tuples in languages other than English we swap BART for a multilingual mT5 and introduce a novel Output Length Regularization to mitigate overfitting to common target sequence lengths, thereby improving the performance of the model by up to 70{\%}. We evaluate our approach on seven datasets in five languages including a zero shot crosslingual setting.}, address = {Seattle, United States}, author = {Pfister, Jan and Wankerl, Sebastian and Hotho, Andreas}, booktitle = {Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)}, keywords = {sentiment}, month = {jul}, pages = {1313–1323}, publisher = {Association for Computational Linguistics}, title = {Point me to your Opinion, {S}en{P}oi}, year = 2022 }
• 
  Semi-unsupervised Learning for Time Series Classification P. Davidson; M. Steininger; A. Huhn; A. Krause; A. Hotho in Milets@KDD (2022).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Time series are ubiquitous and therefore inherently hard to analyze and ultimately to label or cluster. With the rise of the Internet of Things (IoT) and its smart devices, data is collected in large amounts any given second. The collected data is rich in information, as one can detect accidents (e.g. cars) in real time, or assess injury/sickness over a given time span (e.g. health devices). Due to its chaotic nature and massive amounts of datapoints, timeseries are hard to label manually. Furthermore new classes within the data could emerge over time (contrary to e.g. handwritten digits), which would require relabeling the data. In this paper we present SuSL4TS, a deep generative Gaussian mixture model for semi-unsupervised learning, to classify time series data. With our approach we can alleviate manual labeling steps, since we can detect sparsely labeled classes (semi-supervised) and identify emerging classes hidden in the data (unsupervised). We demonstrate the efficacy of our approach with established time series classification datasets from different domains.

  @article{davidson2022semiunsupervised, abstract = {Time series are ubiquitous and therefore inherently hard to analyze and ultimately to label or cluster. With the rise of the Internet of Things (IoT) and its smart devices, data is collected in large amounts any given second. The collected data is rich in information, as one can detect accidents (e.g. cars) in real time, or assess injury/sickness over a given time span (e.g. health devices). Due to its chaotic nature and massive amounts of datapoints, timeseries are hard to label manually. Furthermore new classes within the data could emerge over time (contrary to e.g. handwritten digits), which would require relabeling the data. In this paper we present SuSL4TS, a deep generative Gaussian mixture model for semi-unsupervised learning, to classify time series data. With our approach we can alleviate manual labeling steps, since we can detect sparsely labeled classes (semi-supervised) and identify emerging classes hidden in the data (unsupervised). We demonstrate the efficacy of our approach with established time series classification datasets from different domains.}, author = {Davidson, Padraig and Steininger, Michael and Huhn, André and Krause, Anna and Hotho, Andreas}, journal = {Milets@KDD}, keywords = {timeseries}, publisher = {arXiv}, title = {Semi-unsupervised Learning for Time Series Classification}, year = 2022 }
• 
  Do Different Deep Metric Learning Losses Lead to Similar Learned Features? K. Kobs; M. Steininger; A. Dulny; A. Hotho in CoRR (2022). abs/2205.02698
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{DBLP:journals/corr/abs-2205-02698, author = {Kobs, Konstantin and Steininger, Michael and Dulny, Andrzej and Hotho, Andreas}, journal = {CoRR}, keywords = {myown}, title = {Do Different Deep Metric Learning Losses Lead to Similar Learned Features?}, volume = {abs/2205.02698}, year = 2022 }
• 
  On Learning Hierarchical Embeddings from Encrypted Network Traffic N. Wehner; M. Ring; J. Sch{\"{u}}ler; A. Hotho; T. Ho{\ss}feld; M. Seufert (2022). 1–7.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{DBLP:conf/noms/WehnerRSHHS22, author = {Wehner, Nikolas and Ring, Markus and Sch{\"{u}}ler, Joshua and Hotho, Andreas and Ho{\ss}feld, Tobias and Seufert, Michael}, booktitle = {2022 {IEEE/IFIP} Network Operations and Management Symposium, {NOMS} 2022, Budapest, Hungary, April 25-29, 2022}, keywords = {myown}, pages = {1–7}, publisher = {{IEEE}}, title = {On Learning Hierarchical Embeddings from Encrypted Network Traffic}, year = 2022 }
• 
  SenPoi at SemEval-2022 Task 10: Point me to your Opinion, SenPoi J. Pfister; S. Wankerl; A. Hotho G. Emerson, N. Schluter, G. Stanovsky, R. Kumar, A. Palmer, N. Schneider, S. Singh, S. Ratan (Eds.) (2022). 1313–1323.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{DBLP:conf/semeval/PfisterWH22, author = {Pfister, Jan and Wankerl, Sebastian and Hotho, Andreas}, booktitle = {Proceedings of the 16th International Workshop on Semantic Evaluation, SemEval@NAACL 2022, Seattle, Washington, United States, July 14-15, 2022}, editor = {Emerson, Guy and Schluter, Natalie and Stanovsky, Gabriel and Kumar, Ritesh and Palmer, Alexis and Schneider, Nathan and Singh, Siddharth and Ratan, Shyam}, keywords = {myown}, pages = {1313–1323}, publisher = {Association for Computational Linguistics}, title = {SenPoi at SemEval-2022 Task 10: Point me to your Opinion, SenPoi}, year = 2022 }
• 
  InDiReCT: Language-Guided Zero-Shot Deep Metric Learning for Images K. Kobs; M. Steininger; A. Hotho (2022).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Common Deep Metric Learning (DML) datasets specify only one notion of similarity, e.g., two images in the Cars196 dataset are deemed similar if they show the same car model. We argue that depending on the application, users of image retrieval systems have different and changing similarity notions that should be incorporated as easily as possible. Therefore, we present Language-Guided Zero-Shot Deep Metric Learning (LanZ-DML) as a new DML setting in which users control the properties that should be important for image representations without training data by only using natural language. To this end, we propose InDiReCT (Image representations using Dimensionality Reduction on CLIP embedded Texts), a model for LanZ-DML on images that exclusively uses a few text prompts for training. InDiReCT utilizes CLIP as a fixed feature extractor for images and texts and transfers the variation in text prompt embeddings to the image embedding space. Extensive experiments on five datasets and overall thirteen similarity notions show that, despite not seeing any images during training, InDiReCT performs better than strong baselines and approaches the performance of fully-supervised models. An analysis reveals that InDiReCT learns to focus on regions of the image that correlate with the desired similarity notion, which makes it a fast to train and easy to use method to create custom embedding spaces only using natural language.

  @misc{kobs2022indirect, abstract = {Common Deep Metric Learning (DML) datasets specify only one notion of similarity, e.g., two images in the Cars196 dataset are deemed similar if they show the same car model. We argue that depending on the application, users of image retrieval systems have different and changing similarity notions that should be incorporated as easily as possible. Therefore, we present Language-Guided Zero-Shot Deep Metric Learning (LanZ-DML) as a new DML setting in which users control the properties that should be important for image representations without training data by only using natural language. To this end, we propose InDiReCT (Image representations using Dimensionality Reduction on CLIP embedded Texts), a model for LanZ-DML on images that exclusively uses a few text prompts for training. InDiReCT utilizes CLIP as a fixed feature extractor for images and texts and transfers the variation in text prompt embeddings to the image embedding space. Extensive experiments on five datasets and overall thirteen similarity notions show that, despite not seeing any images during training, InDiReCT performs better than strong baselines and approaches the performance of fully-supervised models. An analysis reveals that InDiReCT learns to focus on regions of the image that correlate with the desired similarity notion, which makes it a fast to train and easy to use method to create custom embedding spaces only using natural language.}, author = {Kobs, Konstantin and Steininger, Michael and Hotho, Andreas}, keywords = {zero}, note = {cite arxiv:2211.12760Comment: Accepted to WACV 2023}, title = {InDiReCT: Language-Guided Zero-Shot Deep Metric Learning for Images}, year = 2022 }
• 
  {W}ue{D}evils at {S}em{E}val-2022 Task 8: Multilingual News Article Similarity via Pair-Wise Sentence Similarity Matrices D. Wangsadirdja; F. Heinickel; S. Trapp; A. Zehe; K. Kobs; A. Hotho (2022). 1235–1243.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  We present a system that creates pair-wise cosine and arccosine sentence similarity matrices using multilingual sentence embeddings obtained from pre-trained SBERT and Universal Sentence Encoder (USE) models respectively. For each news article sentence, it searches the most similar sentence from the other article and computes an average score. Further, a convolutional neural network calculates a total similarity score for the article pairs on these matrices. Finally, a random forest regressor merges the previous results to a final score that can optionally be extended with a publishing date score.

  @inproceedings{wangsadirdja2022wuedevils, abstract = {We present a system that creates pair-wise cosine and arccosine sentence similarity matrices using multilingual sentence embeddings obtained from pre-trained SBERT and Universal Sentence Encoder (USE) models respectively. For each news article sentence, it searches the most similar sentence from the other article and computes an average score. Further, a convolutional neural network calculates a total similarity score for the article pairs on these matrices. Finally, a random forest regressor merges the previous results to a final score that can optionally be extended with a publishing date score.}, address = {Seattle, United States}, author = {Wangsadirdja, Dirk and Heinickel, Felix and Trapp, Simon and Zehe, Albin and Kobs, Konstantin and Hotho, Andreas}, booktitle = {Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)}, keywords = {similarity}, month = {jul}, pages = {1235–1243}, publisher = {Association for Computational Linguistics}, title = {{W}ue{D}evils at {S}em{E}val-2022 Task 8: Multilingual News Article Similarity via Pair-Wise Sentence Similarity Matrices}, year = 2022 }
• 
  {LSX} team5 at {S}em{E}val-2022 Task 8: Multilingual News Article Similarity Assessment based on Word- and Sentence Mover{’}s Distance S. Heil; K. Kopp; A. Zehe; K. Kobs; A. Hotho (2022). 1190–1195.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  This paper introduces our submission for the SemEval 2022 Task 8: Multilingual News Article Similarity. The task of the competition consisted of the development of a model, capable of determining the similarity between pairs of multilingual news articles. To address this challenge, we evaluated the Word Mover{'}s Distance in conjunction with word embeddings from ConceptNet Numberbatch and term frequencies of WorldLex, as well the Sentence Mover{'}s Distance based on sentence embeddings generated by pretrained transformer models of Sentence-BERT. To facilitate the comparison of multilingual articles with Sentence-BERT models, we deployed a Neural Machine Translation system. All our models achieve stable results in multilingual similarity estimation without learning parameters.

  @inproceedings{heil2022lsxteam5, abstract = {This paper introduces our submission for the SemEval 2022 Task 8: Multilingual News Article Similarity. The task of the competition consisted of the development of a model, capable of determining the similarity between pairs of multilingual news articles. To address this challenge, we evaluated the Word Mover{'}s Distance in conjunction with word embeddings from ConceptNet Numberbatch and term frequencies of WorldLex, as well the Sentence Mover{'}s Distance based on sentence embeddings generated by pretrained transformer models of Sentence-BERT. To facilitate the comparison of multilingual articles with Sentence-BERT models, we deployed a Neural Machine Translation system. All our models achieve stable results in multilingual similarity estimation without learning parameters.}, address = {Seattle, United States}, author = {Heil, Stefan and Kopp, Karina and Zehe, Albin and Kobs, Konstantin and Hotho, Andreas}, booktitle = {Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)}, keywords = {similarity}, month = {jul}, pages = {1190–1195}, publisher = {Association for Computational Linguistics}, title = {{LSX} team5 at {S}em{E}val-2022 Task 8: Multilingual News Article Similarity Assessment based on Word- and Sentence Mover{'}s Distance}, year = 2022 }
• 
  Open ERP System Data For Occupational Fraud Detection J. Tritscher; F. Gwinner; D. Schlör; A. Krause; A. Hotho in arxiv (2022).
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{tritscher2022erpData, author = {Tritscher, Julian and Gwinner, Fabian and Schlör, Daniel and Krause, Anna and Hotho, Andreas}, journal = {arxiv}, keywords = {myown}, title = {Open ERP System Data For Occupational Fraud Detection}, year = 2022 }
• 
  NeuralPDE: Modelling Dynamical Systems from Data A. Dulny; A. Hotho; A. Krause in Lecture Notes in Computer Science, R. Bergmann, L. Malburg, S. C. Rodermund, I. J. Timm (Eds.) (2022). (Vol. 13404) 75–89.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{DBLP:conf/ki/DulnyHK22, author = {Dulny, Andrzej and Hotho, Andreas and Krause, Anna}, booktitle = {{KI} 2022: Advances in Artificial Intelligence - 45th German Conference on AI, Trier, Germany, September 19-23, 2022, Proceedings}, editor = {Bergmann, Ralph and Malburg, Lukas and Rodermund, Stephanie C. and Timm, Ingo J.}, keywords = {myown}, pages = {75–89}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {NeuralPDE: Modelling Dynamical Systems from Data}, volume = 13404, year = 2022 }

• 
  A financial game with opportunities for fraud J. Tritscher; A. Krause; D. Schl{\"o}r; F. Gwinner; S. von Mammen; A. Hotho in IEE COG 2021 (2021). 2021
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Even though companies store large amounts of business data in enterprise resource planning (ERP) systems, obtaining data for financial fraud detection is prohibitively difficult due to privacy concerns and companies protecting trade secrets. One possible solution is game-based generation of synthetic ERP data, which poses the challenge of designing an environment that generates realistic ERP data and allows players to commit many different types of fraud. In this work, we design a multiplayer game that allows players to cooperatively run a fictional company, while simultaneously challenging them to maximize their personal gain. We introduce an approach for letting players explore fraud scenarios through emergent gameplay and present a prototype that may be primed with information from real world ERP systems to generate realistic data.

  @article{tritscherfinancial, abstract = {Even though companies store large amounts of business data in enterprise resource planning (ERP) systems, obtaining data for financial fraud detection is prohibitively difficult due to privacy concerns and companies protecting trade secrets. One possible solution is game-based generation of synthetic ERP data, which poses the challenge of designing an environment that generates realistic ERP data and allows players to commit many different types of fraud. In this work, we design a multiplayer game that allows players to cooperatively run a fictional company, while simultaneously challenging them to maximize their personal gain. We introduce an approach for letting players explore fraud scenarios through emergent gameplay and present a prototype that may be primed with information from real world ERP systems to generate realistic data.}, author = {Tritscher, Julian and Krause, Anna and Schl{\"o}r, Daniel and Gwinner, Fabian and von Mammen, Sebastian and Hotho, Andreas}, journal = {IEE COG 2021}, keywords = {myown}, title = {A financial game with opportunities for fraud}, volume = 2021, year = 2021 }
• 
  Proximity dimensions and the emergence of collaboration: a HypTrails study on German AI research T. Koopmann; M. Stubbemann; M. Kapa; M. Paris; G. Buenstorf; T. Hanika; A. Hotho; R. Jäschke; G. Stumme in Scientometrics (2021).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Creation and exchange of knowledge depends on collaboration. Recent work has suggested that the emergence of collaboration frequently relies on geographic proximity. However, being co-located tends to be associated with other dimensions of proximity, such as social ties or a shared organizational environment. To account for such factors, multiple dimensions of proximity have been proposed, including cognitive, institutional, organizational, social and geographical proximity. Since they strongly interrelate, disentangling these dimensions and their respective impact on collaboration is challenging. To address this issue, we propose various methods for measuring different dimensions of proximity. We then present an approach to compare and rank them with respect to the extent to which they indicate co-publications and co-inventions. We adapt the HypTrails approach, which was originally developed to explain human navigation, to co-author and co-inventor graphs. We evaluate this approach on a subset of the German research community, specifically academic authors and inventors active in research on artificial intelligence (AI). We find that social proximity and cognitive proximity are more important for the emergence of collaboration than geographic proximity.

  @article{koopmann2021proximity, abstract = {Creation and exchange of knowledge depends on collaboration. Recent work has suggested that the emergence of collaboration frequently relies on geographic proximity. However, being co-located tends to be associated with other dimensions of proximity, such as social ties or a shared organizational environment. To account for such factors, multiple dimensions of proximity have been proposed, including cognitive, institutional, organizational, social and geographical proximity. Since they strongly interrelate, disentangling these dimensions and their respective impact on collaboration is challenging. To address this issue, we propose various methods for measuring different dimensions of proximity. We then present an approach to compare and rank them with respect to the extent to which they indicate co-publications and co-inventions. We adapt the HypTrails approach, which was originally developed to explain human navigation, to co-author and co-inventor graphs. We evaluate this approach on a subset of the German research community, specifically academic authors and inventors active in research on artificial intelligence (AI). We find that social proximity and cognitive proximity are more important for the emergence of collaboration than geographic proximity.}, author = {Koopmann, Tobias and Stubbemann, Maximilian and Kapa, Matthias and Paris, Michael and Buenstorf, Guido and Hanika, Tom and Hotho, Andreas and Jäschke, Robert and Stumme, Gerd}, journal = {Scientometrics}, keywords = {regio}, title = {Proximity dimensions and the emergence of collaboration: a HypTrails study on German AI research}, year = 2021 }
• 
  Shared Task on Scene Segmentation @ KONVENS 2021 A. Zehe; L. Konle; S. Guhr; L. Dümpelmann; E. Gius; A. Hotho; F. Jannidis; L. Kaufmann; M. Krug; F. Puppe; N. Reiter; A. Schreiber (2021). 1–21.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{zehe2021shared, author = {Zehe, Albin and Konle, Leonard and Guhr, Svenja and Dümpelmann, Lea and Gius, Evelyn and Hotho, Andreas and Jannidis, Fotis and Kaufmann, Lucas and Krug, Markus and Puppe, Frank and Reiter, Nils and Schreiber, Annekea}, booktitle = {Shared Task on Scene Segmentation @ KONVENS 2021}, keywords = {stss}, pages = {1-21}, title = {Shared Task on Scene Segmentation @ KONVENS 2021}, year = 2021 }
• 
  Learning Mathematical Relations Using Deep Tree Models S. Wankerl; A. Dulny; G. Götz; A. Hotho (2021). 1681–1687.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  To the present day, computer algebra systems used for calculating mathematical relations on a symbolic level are mainly rule-based systems. Only recently, deep learning has been applied to the task of symbolic mathematics. Researchers succeeded in developing neural networks which can do symbolic calculations of the equivalence of multiple pairing of equations. However, the generator used to create mathematical terms in previous research has several drawbacks and as with all deep learning tasks, the quality of the data used for training the models has a significant impact on the quality of the results. In this work, we propose a new generator for polynomials. We use it to train several recursive neural networks to recognize the equivalence, derivative, or variable substitution between pairs of polynomials for the first time. Our results indicate that these mathematical relations are identifiable with the help of deep learning.

  @inproceedings{9680206, abstract = {To the present day, computer algebra systems used for calculating mathematical relations on a symbolic level are mainly rule-based systems. Only recently, deep learning has been applied to the task of symbolic mathematics. Researchers succeeded in developing neural networks which can do symbolic calculations of the equivalence of multiple pairing of equations. However, the generator used to create mathematical terms in previous research has several drawbacks and as with all deep learning tasks, the quality of the data used for training the models has a significant impact on the quality of the results. In this work, we propose a new generator for polynomials. We use it to train several recursive neural networks to recognize the equivalence, derivative, or variable substitution between pairs of polynomials for the first time. Our results indicate that these mathematical relations are identifiable with the help of deep learning.}, author = {Wankerl, Sebastian and Dulny, Andrzej and Götz, Gerhard and Hotho, Andreas}, booktitle = {2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA)}, keywords = {treemodels}, month = {dec}, pages = {1681-1687}, title = {Learning Mathematical Relations Using Deep Tree Models}, year = 2021 }
• 
  The Semantic Web - {ISWC} 2021 - 20th International Semantic Web Conference, {ISWC} 2021, Virtual Event, October 24-28, 2021, Proceedings A. Hotho; E. Blomqvist; S. Dietze; A. Fokoue; Y. Ding; P. M. Barnaghi; A. Haller; M. Dragoni; H. Alani in Lecture Notes in Computer Science (2021). (Vol. 12922) Springer.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @proceedings{DBLP:conf/semweb/2021, editor = {Hotho, Andreas and Blomqvist, Eva and Dietze, Stefan and Fokoue, Achille and Ding, Ying and Barnaghi, Payam M. and Haller, Armin and Dragoni, Mauro and Alani, Harith}, keywords = {web}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {The Semantic Web - {ISWC} 2021 - 20th International Semantic Web Conference, {ISWC} 2021, Virtual Event, October 24-28, 2021, Proceedings}, volume = 12922, year = 2021 }
• 
  Malware detection on windows audit logs using LSTMs M. Ring; D. Schlör; S. Wunderlich; D. Landes; A. Hotho in Computers & Security (2021). 109 102389.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Malware is a constant threat and is continuously evolving. Security systems try to keep up with the constant change. One challenge that arises is the large amount of logs generated on an operating system and the need to clarify which information contributes to the detection of possible malware. This work aims at the detection of malware using neural networks based on Windows audit log events. Neural networks can only process continuous data, but Windows audit logs are sequential and textual data. To address these challenges, we extract features out of the audit log events and use LSTMs to capture sequential effects. We create different subsets of features and analyze the effects of additional information. Features describe for example the action-type of windows audit log events, process names or target files that are accessed. Textual features are represented either as one-hot encoding or embedding representation, for which we compare three different approaches for representation learning. Effects of different feature subsets and representations are evaluated on a publicly available data set. Results indicate that using additional information improves the performance of the LSTM-model. While different representations lead to similar classification results, analysis of the latent space shows differences more precisely where FastText seems to be the most promising representation.

  @article{RING2021102389, abstract = {Malware is a constant threat and is continuously evolving. Security systems try to keep up with the constant change. One challenge that arises is the large amount of logs generated on an operating system and the need to clarify which information contributes to the detection of possible malware. This work aims at the detection of malware using neural networks based on Windows audit log events. Neural networks can only process continuous data, but Windows audit logs are sequential and textual data. To address these challenges, we extract features out of the audit log events and use LSTMs to capture sequential effects. We create different subsets of features and analyze the effects of additional information. Features describe for example the action-type of windows audit log events, process names or target files that are accessed. Textual features are represented either as one-hot encoding or embedding representation, for which we compare three different approaches for representation learning. Effects of different feature subsets and representations are evaluated on a publicly available data set. Results indicate that using additional information improves the performance of the LSTM-model. While different representations lead to similar classification results, analysis of the latent space shows differences more precisely where FastText seems to be the most promising representation.}, author = {Ring, Markus and Schlör, Daniel and Wunderlich, Sarah and Landes, Dieter and Hotho, Andreas}, journal = {Computers & Security}, keywords = {myown}, pages = 102389, title = {Malware detection on windows audit logs using LSTMs}, volume = 109, year = 2021 }
• 
  NeuralPDE: Modelling Dynamical Systems from Data A. Dulny; A. Hotho; A. Krause (2021).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Many physical processes such as weather phenomena or fluid mechanics are governed by partial differential equations (PDEs). Modelling such dynamical systems using Neural Networks is an emerging research field. However, current methods are restricted in various ways: they require prior knowledge about the governing equations, and are limited to linear or first-order equations. In this work we propose NeuralPDE, a model which combines convolutional neural networks (CNNs) with differentiable ODE solvers to model dynamical systems. We show that the Method of Lines used in standard PDE solvers can be represented using convolutions which makes CNNs the natural choice to parametrize arbitrary PDE dynamics. Our model can be applied to any data without requiring any prior knowledge about the governing PDE. We evaluate NeuralPDE on datasets generated by solving a wide variety of PDEs, covering higher orders, non-linear equations and multiple spatial dimensions.

  @misc{dulny2021neuralpde, abstract = {Many physical processes such as weather phenomena or fluid mechanics are governed by partial differential equations (PDEs). Modelling such dynamical systems using Neural Networks is an emerging research field. However, current methods are restricted in various ways: they require prior knowledge about the governing equations, and are limited to linear or first-order equations. In this work we propose NeuralPDE, a model which combines convolutional neural networks (CNNs) with differentiable ODE solvers to model dynamical systems. We show that the Method of Lines used in standard PDE solvers can be represented using convolutions which makes CNNs the natural choice to parametrize arbitrary PDE dynamics. Our model can be applied to any data without requiring any prior knowledge about the governing PDE. We evaluate NeuralPDE on datasets generated by solving a wide variety of PDEs, covering higher orders, non-linear equations and multiple spatial dimensions.}, author = {Dulny, Andrzej and Hotho, Andreas and Krause, Anna}, keywords = {pde}, note = {cite arxiv:2111.07671}, title = {NeuralPDE: Modelling Dynamical Systems from Data}, year = 2021 }
• 
  Sequential Item Recommendation in the MOBA Game Dota 2 A. Dallmann; J. Kohlmann; D. Zoller; A. Hotho (2021). 10–17.
  • [ Abstract ]
  • [ BibTeX ]
  • [ DOI ]
  Multiplayer Online Battle Arena (MOBA) games such as Dota 2 attract hundreds of thousands of players every year. Despite the large player base, it is still important to attract new players to prevent the community of a game from becoming inactive. Entering MOBA games is, however, often demanding, requiring the player to learn numerous skills at once. An important factor of success is buying the correct items which forms a complex task depending on various in-game factors such as already purchased items, the team composition, or available resources. A recommendation system can support players by reducing the mental effort required to choose a suitable item, helping, e.g., newer players or players returning to the game after a longer break, to focus on other aspects of the game. Since Sequential Item Recommendation (SIR) has proven to be effective in various domains (e.g. e-commerce, movie recommendation or playlist continuation), we explore the applicability of well-known SIR models in the context of purchase recommendations in Dota 2. To facilitate this research, we collect, analyze and publish Dota-350k, a new large dataset based on recent Dota 2 matches. We find that SIR models can be employed effectively for item recommendation in Dota 2. Our results show that models that consider the order of purchases are the most effective. In contrast to other domains, we find RNN-based models to outperform the more recent Transformer-based architectures on Dota-350k.

  @inproceedings{9679858, abstract = {Multiplayer Online Battle Arena (MOBA) games such as Dota 2 attract hundreds of thousands of players every year. Despite the large player base, it is still important to attract new players to prevent the community of a game from becoming inactive. Entering MOBA games is, however, often demanding, requiring the player to learn numerous skills at once. An important factor of success is buying the correct items which forms a complex task depending on various in-game factors such as already purchased items, the team composition, or available resources. A recommendation system can support players by reducing the mental effort required to choose a suitable item, helping, e.g., newer players or players returning to the game after a longer break, to focus on other aspects of the game. Since Sequential Item Recommendation (SIR) has proven to be effective in various domains (e.g. e-commerce, movie recommendation or playlist continuation), we explore the applicability of well-known SIR models in the context of purchase recommendations in Dota 2. To facilitate this research, we collect, analyze and publish Dota-350k, a new large dataset based on recent Dota 2 matches. We find that SIR models can be employed effectively for item recommendation in Dota 2. Our results show that models that consider the order of purchases are the most effective. In contrast to other domains, we find RNN-based models to outperform the more recent Transformer-based architectures on Dota-350k.}, author = {Dallmann, Alexander and Kohlmann, Johannes and Zoller, Daniel and Hotho, Andreas}, booktitle = {2021 International Conference on Data Mining Workshops (ICDMW)}, keywords = {recommender}, month = {dec}, pages = {10-17}, title = {Sequential Item Recommendation in the MOBA Game Dota 2}, year = 2021 }
• 
  CoBERT: Scientific Collaboration Prediction via Sequential Recommendation T. Koopmann; K. Kobs; K. Herud; A. Hotho (2021). 45–54.
  • [ Abstract ]
  • [ BibTeX ]
  • [ DOI ]
  • [ Download ]
  Collaborations are an Important factor for scientific success, as the joint work leads to results individual scientists cannot easily reach. Recommending collaborations automatically can alleviate the time consuming and tedious search for potential collaborators. Usually, such recommendation systems rely on graph structures modeling co-authorship of papers and content-based relations such as similar paper keywords. Models are then trained to estimate the probability of links between certain authors in these graphs.In this paper, we argue that the order of papers is crucial for reliably predicting future collaborations, which is not considered by graph-based recommendation systems. We thus propose to reformulate the task of collaboration recommendation as a sequential recommendation task. Here, we aim to predict the next co-author in a chronologically sorted sequence of an author’s collaborators. We introduce CoBERT, a BERT4Rec inspired model, that predicts the sequence’s next co-author and thus a potential collaborator. Since the order of co-authors of a single paper is not that important compared to the overall paper order, we leverage positional embeddings encoding paper positions instead of co-author positions in the sequence. Additionally, we inject content features about every paper and their co-authors. We evaluate CoBERT on two datasets consisting of papers from the field of Artificial Intelligence and the journal PlosOne. We show that CoBERT can outperform graph-based methods and BERT4Rec when predicting the co-authors of the next paper. We make our code and data available.

  @inproceedings{koopmann2021cobert, abstract = {Collaborations are an Important factor for scientific success, as the joint work leads to results individual scientists cannot easily reach. Recommending collaborations automatically can alleviate the time consuming and tedious search for potential collaborators. Usually, such recommendation systems rely on graph structures modeling co-authorship of papers and content-based relations such as similar paper keywords. Models are then trained to estimate the probability of links between certain authors in these graphs.In this paper, we argue that the order of papers is crucial for reliably predicting future collaborations, which is not considered by graph-based recommendation systems. We thus propose to reformulate the task of collaboration recommendation as a sequential recommendation task. Here, we aim to predict the next co-author in a chronologically sorted sequence of an author’s collaborators. We introduce CoBERT, a BERT4Rec inspired model, that predicts the sequence’s next co-author and thus a potential collaborator. Since the order of co-authors of a single paper is not that important compared to the overall paper order, we leverage positional embeddings encoding paper positions instead of co-author positions in the sequence. Additionally, we inject content features about every paper and their co-authors. We evaluate CoBERT on two datasets consisting of papers from the field of Artificial Intelligence and the journal PlosOne. We show that CoBERT can outperform graph-based methods and BERT4Rec when predicting the co-authors of the next paper. We make our code and data available.}, author = {Koopmann, Tobias and Kobs, Konstantin and Herud, Konstantin and Hotho, Andreas}, booktitle = {2021 International Conference on Data Mining Workshops (ICDMW)}, keywords = {recommender}, month = {dec}, pages = {45-54}, title = {CoBERT: Scientific Collaboration Prediction via Sequential Recommendation}, year = 2021 }
• 
  Detecting Presence Of Speech In Acoustic Data Obtained From Beehives P. Janetzky; P. Davidson; M. Steininger; A. Krause; A. Hotho F. Font, A. Mesaros, D. P. W. Ellis, E. Fonseca, M. Fuentes, B. Elizalde (Eds.) (2021). 26–30.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{janetzky2021detecting, author = {Janetzky, Pascal and Davidson, Padraig and Steininger, Michael and Krause, Anna and Hotho, Andreas}, booktitle = {Proceedings of the 6th Workshop on Detection and Classification of Acoustic Scenes and Events 2021 {(DCASE} 2021), Online, November 15-19, 2021}, editor = {Font, Frederic and Mesaros, Annamaria and Ellis, Daniel P. W. and Fonseca, Eduardo and Fuentes, Magdalena and Elizalde, Benjamin}, keywords = {we4bee}, pages = {26–30}, title = {Detecting Presence Of Speech In Acoustic Data Obtained From Beehives}, year = 2021 }
• 
  Anomaly Detection in Beehives: An Algorithm Comparison P. Davidson; M. Steininger; F. Lautenschlager; A. Krause; A. Hotho (2021).
  • [ BibTeX ]
  @misc{davidson2021anomaly, author = {Davidson, Padraig and Steininger, Michael and Lautenschlager, Florian and Krause, Anna and Hotho, Andreas}, keywords = {we4bee}, title = {Anomaly Detection in Beehives: An Algorithm Comparison}, year = 2021 }
• 
  Self-Supervised Multi-Task Pretraining Improves Image Aesthetic Assessment J. Pfister; K. Kobs; A. Hotho (2021). 816–825.
  • [ BibTeX ]
  @inproceedings{Pfister_2021_CVPR, author = {Pfister, Jan and Kobs, Konstantin and Hotho, Andreas}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, keywords = {myown}, month = {June}, pages = {816-825}, title = {Self-Supervised Multi-Task Pretraining Improves Image Aesthetic Assessment}, year = 2021 }
• 
  Evaluating the multi-task learning approach for land use regression modelling of air pollution A. Dulny; M. Steininger; F. Lautenschlager; A. Krause; A. Hotho in Journal of Physics: Conference Series (2021). 1834(1) 012004.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Air pollution has been linked to several health problems including heart disease, stroke and lung cancer. Modelling and analyzing this dependency requires reliable and accurate air pollutant measurements collected by stationary air monitoring stations. However, usually only a low number of such stations are present within a single city. To retrieve pollution concentrations for unmeasured locations, researchers rely on land use regression (LUR) models. Those models are typically developed for one pollutant only. However, as results in different areas have shown, modelling several related output variables through multi-task learning can improve the prediction results of the models significantly. In this work, we compared prediction results from single-task and multi-task learning multilayer perceptron models on measurements taken from the OpenSense dataset and the London Atmospheric Emissions Inventory dataset. LUR features were generated from OpenStreetMap using OpenLUR and used to train hard parameter sharing multilayer perceptron models. The results show multi-task learning with sufficient data significantly improves the performance of a LUR model.

  @article{dulny2021evaluating, abstract = {Air pollution has been linked to several health problems including heart disease, stroke and lung cancer. Modelling and analyzing this dependency requires reliable and accurate air pollutant measurements collected by stationary air monitoring stations. However, usually only a low number of such stations are present within a single city. To retrieve pollution concentrations for unmeasured locations, researchers rely on land use regression (LUR) models. Those models are typically developed for one pollutant only. However, as results in different areas have shown, modelling several related output variables through multi-task learning can improve the prediction results of the models significantly. In this work, we compared prediction results from single-task and multi-task learning multilayer perceptron models on measurements taken from the OpenSense dataset and the London Atmospheric Emissions Inventory dataset. LUR features were generated from OpenStreetMap using OpenLUR and used to train hard parameter sharing multilayer perceptron models. The results show multi-task learning with sufficient data significantly improves the performance of a LUR model.}, author = {Dulny, Andrzej and Steininger, Michael and Lautenschlager, Florian and Krause, Anna and Hotho, Andreas}, journal = {Journal of Physics: Conference Series}, keywords = {myown}, month = {mar}, number = 1, organization = {IOP Publishing}, pages = {012004}, publisher = {{IOP} Publishing}, title = {Evaluating the multi-task learning approach for land use regression modelling of air pollution}, volume = 1834, year = 2021 }
• 
  Do Different Deep Metric Learning Losses Lead to Similar Learned Features? K. Kobs; M. Steininger; A. Dulny; A. Hotho in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2021). 10644–10654.
  • [ BibTeX ]
  @article{noauthororeditor, author = {Kobs, Konstantin and Steininger, Michael and Dulny, Andrzej and Hotho, Andreas}, journal = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)}, keywords = {myown}, month = {October}, pages = {10644-10654}, title = {Do Different Deep Metric Learning Losses Lead to Similar Learned Features?}, year = 2021 }
• 
  Detecting Scenes in Fiction: A new Segmentation Task A. Zehe; L. Konle; L. Dümpelmann; E. Gius; A. Hotho; F. Jannidis; L. Kaufmann; M. Krug; F. Puppe; N. Reiter; A. Schreiber; N. Wiedmer (2021).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{zehe2021detecting, author = {Zehe, Albin and Konle, Leonard and Dümpelmann, Lea and Gius, Evelyn and Hotho, Andreas and Jannidis, Fotis and Kaufmann, Lucas and Krug, Markus and Puppe, Frank and Reiter, Nils and Schreiber, Annekea and Wiedmer, Nathalie}, booktitle = {Proceedings of the 16th Conference of the {E}uropean Chapter of the Association for Computational Linguistics: Volume 1, Long Papers}, keywords = {task}, publisher = {ACL}, title = {Detecting Scenes in Fiction: A new Segmentation Task}, year = 2021 }
• 
  Comparison of Transformer-Based Sequential Product Recommendation Models for the Coveo Data Challenge E. Fischer; D. Zoller; A. Hotho in SIGIR Workshop On eCommerce (2021).
  • [ BibTeX ]
  • [ Download ]
  @article{fischer2021comparison, author = {Fischer, Elisabeth and Zoller, Daniel and Hotho, Andreas}, journal = {SIGIR Workshop On eCommerce}, keywords = {transformer}, month = {July}, title = {Comparison of Transformer-Based Sequential Product Recommendation Models for the Coveo Data Challenge}, year = 2021 }
• 
  Density-based weighting for imbalanced regression M. Steininger; K. Kobs; P. Davidson; A. Krause; A. Hotho in Machine Learning, (A. Appice; S. Escalera; J. A. Gamez; H. Trautmann, Eds.) (2021).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  In many real world settings,imbalanced data impedes model performance of learning algorithms, like neural networks, mostly for rare cases. This is especially problematic for tasks focusing on these rare occurrences. For example, when estimating precipitation, extreme rainfall events are scarce but important considering their potential consequences. While there are numerous well studied solutions for classification settings, most of them cannot be applied to regression easily. Of the few solutions for regression tasks, barely any have explored cost-sensitive learning which is known to have advantages compared to sampling-based methods in classification tasks. In this work, we propose a sample weighting approach for imbalanced regression datasets called DenseWeight and a cost-sensitive learning approach for neural network regression with imbalanced data called DenseLoss based on our weighting scheme. DenseWeight weights data points according to their target value rarities through kernel density estimation (KDE). DenseLoss adjusts each data point’s influence on the loss according to DenseWeight, giving rare data points more influence on modeltraining compared to common data points. We show on multiple differently distributed datasets that DenseLoss significantly improves model performance for rare data points through its density-based weighting scheme. Additionally, we compare DenseLoss to the state-of-the-art method SMOGN, finding that our method mostly yields better performance. Our approach provides more control over model training as it enables us to actively decide on the trade-off between focusing on common or rare cases through a single hyperparameter, allowing the training of better models for rare data points.

  @article{steininger2021densitybased, abstract = {In many real world settings,imbalanced data impedes model performance of learning algorithms, like neural networks, mostly for rare cases. This is especially problematic for tasks focusing on these rare occurrences. For example, when estimating precipitation, extreme rainfall events are scarce but important considering their potential consequences. While there are numerous well studied solutions for classification settings, most of them cannot be applied to regression easily. Of the few solutions for regression tasks, barely any have explored cost-sensitive learning which is known to have advantages compared to sampling-based methods in classification tasks. In this work, we propose a sample weighting approach for imbalanced regression datasets called DenseWeight and a cost-sensitive learning approach for neural network regression with imbalanced data called DenseLoss based on our weighting scheme. DenseWeight weights data points according to their target value rarities through kernel density estimation (KDE). DenseLoss adjusts each data point’s influence on the loss according to DenseWeight, giving rare data points more influence on modeltraining compared to common data points. We show on multiple differently distributed datasets that DenseLoss significantly improves model performance for rare data points through its density-based weighting scheme. Additionally, we compare DenseLoss to the state-of-the-art method SMOGN, finding that our method mostly yields better performance. Our approach provides more control over model training as it enables us to actively decide on the trade-off between focusing on common or rare cases through a single hyperparameter, allowing the training of better models for rare data points.}, author = {Steininger, Michael and Kobs, Konstantin and Davidson, Padraig and Krause, Anna and Hotho, Andreas}, editor = {Appice, Annalisa and Escalera, Sergio and Gamez, Jose A. and Trautmann, Heike}, journal = {Machine Learning}, keywords = {selected}, title = {Density-based weighting for imbalanced regression}, year = 2021 }
• 
  Semi-unsupervised Learning: An In-depth Parameter Analysis P. Davidson; F. Buckermann; M. Steininger; A. Krause; A. Hotho S. Edelkamp, R. M{\"o}ller, E. Rueckert (Eds.) (2021). 51–66.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Creating datasets for supervised learning is a very challenging and expensive task, in which each input example has to be annotated with its expected output (e.g. object class). By combining unsupervised and semi-supervised learning, semi-unsupervised learning proposes a new paradigm for partially labeled datasets with additional unknown classes. In this paper we focus on a better understanding of this new learning paradigm and analyze the impact of the amount of labeled data, the number of augmented classes and the selection of hidden classes on the quality of prediction. Especially the number of augmented classes highly influences classification accuracy, which needs tuning for each dataset, since too few and too many augmented classes are detrimental to classifier performance. We also show that we can improve results on a large variety of datasets when using convolutional networks as feature extractors while applying output driven entropy regularization instead of a simple weight based L2 norm.

  @inproceedings{10.1007/978-3-030-87626-5_5, abstract = {Creating datasets for supervised learning is a very challenging and expensive task, in which each input example has to be annotated with its expected output (e.g. object class). By combining unsupervised and semi-supervised learning, semi-unsupervised learning proposes a new paradigm for partially labeled datasets with additional unknown classes. In this paper we focus on a better understanding of this new learning paradigm and analyze the impact of the amount of labeled data, the number of augmented classes and the selection of hidden classes on the quality of prediction. Especially the number of augmented classes highly influences classification accuracy, which needs tuning for each dataset, since too few and too many augmented classes are detrimental to classifier performance. We also show that we can improve results on a large variety of datasets when using convolutional networks as feature extractors while applying output driven entropy regularization instead of a simple weight based L2 norm.}, address = {Cham}, author = {Davidson, Padraig and Buckermann, Florian and Steininger, Michael and Krause, Anna and Hotho, Andreas}, booktitle = {KI 2021: Advances in Artificial Intelligence}, editor = {Edelkamp, Stefan and M{\"o}ller, Ralf and Rueckert, Elmar}, keywords = {unsupervised}, pages = {51–66}, publisher = {Springer International Publishing}, title = {Semi-unsupervised Learning: An In-depth Parameter Analysis}, year = 2021 }
• 
  A Case Study on Sampling Strategies for Evaluating Neural Sequential Item Recommendation Models A. Dallmann; D. Zoller; A. Hotho (2021).
  • [ BibTeX ]
  • [ DOI ]
  • [ Download ]
  @inproceedings{dallmann2021study, author = {Dallmann, Alexander and Zoller, Daniel and Hotho, Andreas}, booktitle = {Fifteenth {ACM} Conference on Recommender Systems}, keywords = {systems}, month = {sep}, publisher = {{ACM}}, title = {A Case Study on Sampling Strategies for Evaluating Neural Sequential Item Recommendation Models}, year = 2021 }
• 
  Assessing Media Bias in Cross-Linguistic and Cross-National Populations A. Sales; A. Zehe; L. B. Marinho; A. Veloso; A. Hotho; J. Omeliyanenko in Proceedings of the International AAAI Conference on Web and Social Media (2021). 15(1) 561–572.
  • [ BibTeX ]
  • [ URL ]
  @article{sales2021assessing, author = {Sales, Allan and Zehe, Albin and Marinho, Leandro Balby and Veloso, Adriano and Hotho, Andreas and Omeliyanenko, Janna}, journal = {Proceedings of the International AAAI Conference on Web and Social Media}, keywords = {sentiment}, month = {may}, number = 1, pages = {561-572}, title = {Assessing Media Bias in Cross-Linguistic and Cross-National Populations}, volume = 15, year = 2021 }
• 
  Semi-Supervised Learning for Grain Size Distribution Interpolation K. Kobs; C. Schäfer; M. Steininger; A. Krause; R. Baumhauer; H. Paeth; A. Hotho (2021). 34–44.
  • [ BibTeX ]
  @inproceedings{kobs2021semi, author = {Kobs, Konstantin and Schäfer, Christian and Steininger, Michael and Krause, Anna and Baumhauer, Roland and Paeth, Heiko and Hotho, Andreas}, booktitle = {Pattern Recognition. ICPR International Workshops and Challenges: Virtual Event, January 10–15, 2021, Proceedings, Part VI}, keywords = {myown}, organization = {Springer International Publishing}, pages = {34–44}, title = {Semi-Supervised Learning for Grain Size Distribution Interpolation}, year = 2021 }
• 
  A financial game with opportunities for fraud J. Tritscher; A. Krause; D. Schl{\"{o}}r; F. Gwinner; S. von Mammen; A. Hotho (2021). 1–5.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{DBLP:conf/cig/TritscherKSGMH21, author = {Tritscher, Julian and Krause, Anna and Schl{\"{o}}r, Daniel and Gwinner, Fabian and von Mammen, Sebastian and Hotho, Andreas}, booktitle = {2021 {IEEE} Conference on Games (CoG), Copenhagen, Denmark, August 17-20, 2021}, keywords = {myown}, pages = {1–5}, publisher = {{IEEE}}, title = {A financial game with opportunities for fraud}, year = 2021 }

• 
  Where to Submit? Helping Researchers to Choose the Right Venue K. Kobs; T. Koopmann; A. Zehe; D. Fernes; P. Krop; A. Hotho (2020). 878–883.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Whenever researchers write a paper, the same question occurs: {``}Where to submit?{''} In this work, we introduce WTS, an open and interpretable NLP system that recommends conferences and journals to researchers based on the title, abstract, and/or keywords of a given paper. We adapt the TextCNN architecture and automatically analyze its predictions using the Integrated Gradients method to highlight words and phrases that led to the recommendation of a scientific venue. We train and test our method on publications from the fields of artificial intelligence (AI) and medicine, both derived from the Semantic Scholar dataset. WTS achieves an Accuracy@5 of approximately 83{\%} for AI papers and 95{\%} in the field of medicine. It is open source and available for testing on https://wheretosubmit.ml.

  @inproceedings{kobs2020where, abstract = {Whenever researchers write a paper, the same question occurs: {``}Where to submit?{''} In this work, we introduce WTS, an open and interpretable NLP system that recommends conferences and journals to researchers based on the title, abstract, and/or keywords of a given paper. We adapt the TextCNN architecture and automatically analyze its predictions using the Integrated Gradients method to highlight words and phrases that led to the recommendation of a scientific venue. We train and test our method on publications from the fields of artificial intelligence (AI) and medicine, both derived from the Semantic Scholar dataset. WTS achieves an Accuracy@5 of approximately 83{\%} for AI papers and 95{\%} in the field of medicine. It is open source and available for testing on https://wheretosubmit.ml.}, address = {Online}, author = {Kobs, Konstantin and Koopmann, Tobias and Zehe, Albin and Fernes, David and Krop, Philipp and Hotho, Andreas}, booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2020}, keywords = {venue}, month = {nov}, pages = {878–883}, publisher = {Association for Computational Linguistics}, title = {Where to Submit? Helping Researchers to Choose the Right Venue}, year = 2020 }
• 
  Deep Learning for Climate Model Output Statistics M. Steininger; D. Abel; K. Ziegler; A. Krause; H. Paeth; A. Hotho in Tackling Climate Change with Machine Learning Workshop at NeurIPS 2020 (2020).
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @article{steininger2020learning, author = {Steininger, Michael and Abel, Daniel and Ziegler, Katrin and Krause, Anna and Paeth, Heiko and Hotho, Andreas}, journal = {Tackling Climate Change with Machine Learning Workshop at NeurIPS 2020}, keywords = {myown}, title = {Deep Learning for Climate Model Output Statistics}, year = 2020 }
• 
  Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part III U. Brefeld; Élisa Fromont; A. Hotho; A. J. Knobbe; M. H. Maathuis; C. Robardet in Lecture Notes in Computer Science (2020). (Vol. 11908) Springer.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{brefeld2020machine1, booktitle = {ECML/PKDD (3)}, editor = {Brefeld, Ulf and Fromont, Élisa and Hotho, Andreas and Knobbe, Arno J. and Maathuis, Marloes H. and Robardet, Céline}, keywords = {proceedings}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part III}, volume = 11908, year = 2020 }
• 
  Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part II U. Brefeld; Élisa Fromont; A. Hotho; A. J. Knobbe; M. H. Maathuis; C. Robardet in Lecture Notes in Computer Science (2020). (Vol. 11907) Springer.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{brefeld2020machine2, booktitle = {ECML/PKDD (2)}, editor = {Brefeld, Ulf and Fromont, Élisa and Hotho, Andreas and Knobbe, Arno J. and Maathuis, Marloes H. and Robardet, Céline}, keywords = {proceedings}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part II}, volume = 11907, year = 2020 }
• 
  Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part I U. Brefeld; Élisa Fromont; A. Hotho; A. J. Knobbe; M. H. Maathuis; C. Robardet in Lecture Notes in Computer Science (2020). (Vol. 11906) Springer.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{brefeld2020machine3, booktitle = {ECML/PKDD (1)}, editor = {Brefeld, Ulf and Fromont, Élisa and Hotho, Andreas and Knobbe, Arno J. and Maathuis, Marloes H. and Robardet, Céline}, keywords = {proceedings}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part I}, volume = 11906, year = 2020 }
• 
  SimLoss: Class Similarities in Cross Entropy K. Kobs; M. Steininger; A. Zehe; F. Lautenschlager; A. Hotho D. Helic, G. Leitner, M. Stettinger, A. Felfernig, Z. W. Ra{\’{s}} (Eds.) (2020). 431–439.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  One common loss function in neural network classification tasks is Categorical Cross Entropy (CCE), which punishes all misclassifications equally. However, classes often have an inherent structure. For instance, classifying an image of a rose as ``violet'' is better than as ``truck''. We introduce SimLoss, a drop-in replacement for CCE that incorporates class similarities along with two techniques to construct such matrices from task-specific knowledge. We test SimLoss on Age Estimation and Image Classification and find that it brings significant improvements over CCE on several metrics. SimLoss therefore allows for explicit modeling of background knowledge by simply exchanging the loss function, while keeping the neural network architecture the same. Code and additional resources are available at https://github.com/konstantinkobs/SimLoss

  @inproceedings{kobs2020simloss, abstract = {One common loss function in neural network classification tasks is Categorical Cross Entropy (CCE), which punishes all misclassifications equally. However, classes often have an inherent structure. For instance, classifying an image of a rose as ``violet'' is better than as ``truck''. We introduce SimLoss, a drop-in replacement for CCE that incorporates class similarities along with two techniques to construct such matrices from task-specific knowledge. We test SimLoss on Age Estimation and Image Classification and find that it brings significant improvements over CCE on several metrics. SimLoss therefore allows for explicit modeling of background knowledge by simply exchanging the loss function, while keeping the neural network architecture the same. Code and additional resources are available at https://github.com/konstantinkobs/SimLoss}, address = {Cham}, author = {Kobs, Konstantin and Steininger, Michael and Zehe, Albin and Lautenschlager, Florian and Hotho, Andreas}, booktitle = {Foundations of Intelligent Systems}, editor = {Helic, Denis and Leitner, Gerhard and Stettinger, Martin and Felfernig, Alexander and Ra{\'{s}}, Zbigniew W.}, keywords = {myown}, pages = {431–439}, publisher = {Springer International Publishing}, title = {SimLoss: Class Similarities in Cross Entropy}, year = 2020 }
• 
  LM4KG: Improving Common Sense Knowledge Graphs with Language Models J. Omeliyanenko; A. Zehe; L. Hettinger; A. Hotho J. Z. Pan, V. Tamma, C. d’Amato, K. Janowicz, B. Fu, A. Polleres, O. Seneviratne, L. Kagal (Eds.) (2020). 456–473.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Language Models (LMs) and Knowledge Graphs (KGs) are both active research areas in Machine Learning and Semantic Web. While LMs have brought great improvements for many downstream tasks on their own, they are often combined with KGs providing additionally aggregated, well structured knowledge. Usually, this is done by leveraging KGs to improve LMs. But what happens if we turn this around and use LMs to improve KGs?

  @inproceedings{omeliyanenko2020lm4kg, abstract = {Language Models (LMs) and Knowledge Graphs (KGs) are both active research areas in Machine Learning and Semantic Web. While LMs have brought great improvements for many downstream tasks on their own, they are often combined with KGs providing additionally aggregated, well structured knowledge. Usually, this is done by leveraging KGs to improve LMs. But what happens if we turn this around and use LMs to improve KGs?}, address = {Cham}, author = {Omeliyanenko, Janna and Zehe, Albin and Hettinger, Lena and Hotho, Andreas}, booktitle = {The Semantic Web – ISWC 2020}, editor = {Pan, Jeff Z. and Tamma, Valentina and d'Amato, Claudia and Janowicz, Krzysztof and Fu, Bo and Polleres, Axel and Seneviratne, Oshani and Kagal, Lalana}, keywords = {semantic}, pages = {456–473}, publisher = {Springer International Publishing}, title = {LM4KG: Improving Common Sense Knowledge Graphs with Language Models}, year = 2020 }
• 
  Evaluation of Post-hoc XAI Approaches Through Synthetic Tabular Data J. Tritscher; M. Ring; D. Schlr; L. Hettinger; A. Hotho D. Helic, G. Leitner, M. Stettinger, A. Felfernig, Z. W. Ra{\’{s}} (Eds.) (2020). 422–430.
  • [ Abstract ]
  • [ BibTeX ]
  • [ Download ]
  Evaluating the explanations given by post-hoc XAI approaches on tabular data is a challenging prospect, since the subjective judgement of explanations of tabular relations is non trivial in contrast to e.g. the judgement of image heatmap explanations. In order to quantify XAI performance on categorical tabular data, where feature relationships can often be described by Boolean functions, we propose an evaluation setting through generation of synthetic datasets. To create gold standard explanations, we present a definition of feature relevance in Boolean functions. In the proposed setting we evaluate eight state-of-the-art XAI approaches and gain novel insights into XAI performance on categorical tabular data. We find that the investigated approaches often fail to faithfully explain even basic relationships within categorical data.

  @inproceedings{tritscher2020evaluation, abstract = {Evaluating the explanations given by post-hoc XAI approaches on tabular data is a challenging prospect, since the subjective judgement of explanations of tabular relations is non trivial in contrast to e.g. the judgement of image heatmap explanations. In order to quantify XAI performance on categorical tabular data, where feature relationships can often be described by Boolean functions, we propose an evaluation setting through generation of synthetic datasets. To create gold standard explanations, we present a definition of feature relevance in Boolean functions. In the proposed setting we evaluate eight state-of-the-art XAI approaches and gain novel insights into XAI performance on categorical tabular data. We find that the investigated approaches often fail to faithfully explain even basic relationships within categorical data.}, address = {Cham}, author = {Tritscher, Julian and Ring, Markus and Schlr, Daniel and Hettinger, Lena and Hotho, Andreas}, booktitle = {Foundations of Intelligent Systems}, editor = {Helic, Denis and Leitner, Gerhard and Stettinger, Martin and Felfernig, Alexander and Ra{\'{s}}, Zbigniew W.}, keywords = {xai}, pages = {422–430}, publisher = {Springer International Publishing}, title = {Evaluation of Post-hoc XAI Approaches Through Synthetic Tabular Data}, year = 2020 }
• 
  OpenLUR: Off-the-shelf air pollution modeling with open features and machine learning F. Lautenschlager; M. Becker; K. Kobs; M. Steininger; P. Davidson; A. Krause; A. Hotho in Atmospheric Environment (2020). 233 117535.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  To assess the exposure of citizens to pollutants like NOx or particulate matter in urban areas, land use regression (LUR) models are a well established method. LUR models leverage information about environmental and anthropogenic factors such as cars, heating, or industry to predict air pollution in areas where no measurements have been made. However, existing approaches are often not globally applicable and require tedious hyper-parameter tuning to enable high quality predictions. In this work, we tackle these issues by introducing OpenLUR, an off-the-shelf approach for modeling air pollution that (i) works on a set of novel features solely extracted from the globally and openly available data source OpenStreetMap and (ii) is based on state-of-the-art machine learning featuring automated hyper-parameter tuning in order to minimize manual effort. We show that our proposed features are able to outperform their counterparts from local and closed sources, and illustrate how automated hyper parameter tuning can yield competitve results while alleviating the need for expert knowledge in machine learning and manual effort. Importantly, we further demonstrate the potential of the global availability of our features by applying cross-learning across different cities in order to reduce the need for a large amount of training samples. Overall, OpenLUR represents an off-the-shelf approach that facilitates easily reproducible experiments and the development of globally applicable models.

  @article{lautenschlager2020openlur, abstract = {To assess the exposure of citizens to pollutants like NOx or particulate matter in urban areas, land use regression (LUR) models are a well established method. LUR models leverage information about environmental and anthropogenic factors such as cars, heating, or industry to predict air pollution in areas where no measurements have been made. However, existing approaches are often not globally applicable and require tedious hyper-parameter tuning to enable high quality predictions. In this work, we tackle these issues by introducing OpenLUR, an off-the-shelf approach for modeling air pollution that (i) works on a set of novel features solely extracted from the globally and openly available data source OpenStreetMap and (ii) is based on state-of-the-art machine learning featuring automated hyper-parameter tuning in order to minimize manual effort. We show that our proposed features are able to outperform their counterparts from local and closed sources, and illustrate how automated hyper parameter tuning can yield competitve results while alleviating the need for expert knowledge in machine learning and manual effort. Importantly, we further demonstrate the potential of the global availability of our features by applying cross-learning across different cities in order to reduce the need for a large amount of training samples. Overall, OpenLUR represents an off-the-shelf approach that facilitates easily reproducible experiments and the development of globally applicable models.}, author = {Lautenschlager, Florian and Becker, Martin and Kobs, Konstantin and Steininger, Michael and Davidson, Padraig and Krause, Anna and Hotho, Andreas}, journal = {Atmospheric Environment}, keywords = {pollution}, pages = 117535, title = {OpenLUR: Off-the-shelf air pollution modeling with open features and machine learning}, volume = 233, year = 2020 }
• 
  MapLUR: Exploring a New Paradigm for Estimating Air Pollution Using Deep Learning on Map Images M. Steininger; K. Kobs; A. Zehe; F. Lautenschlager; M. Becker; A. Hotho in ACM Trans. Spatial Algorithms Syst. (2020). 6(3)
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Land-use regression (LUR) models are important for the assessment of air pollution concentrations in areas without measurement stations. While many such models exist, they often use manually constructed features based on restricted, locally available data. Thus, they are typically hard to reproduce and challenging to adapt to areas beyond those they have been developed for. In this paper, we advocate a paradigm shift for LUR models: We propose the Data-driven, Open, Global (DOG) paradigm that entails models based on purely data-driven approaches using only openly and globally available data. Progress within this paradigm will alleviate the need for experts to adapt models to the local characteristics of the available data sources and thus facilitate the generalizability of air pollution models to new areas on a global scale. In order to illustrate the feasibility of the DOG paradigm for LUR, we introduce a deep learning model called MapLUR. It is based on a convolutional neural network architecture and is trained exclusively on globally and openly available map data without requiring manual feature engineering. We compare our model to state-of-the-art baselines like linear regression, random forests and multi-layer perceptrons using a large data set of modeled NO2 concentrations in Central London. Our results show that MapLUR significantly outperforms these approaches even though they are provided with manually tailored features. Furthermore, we illustrate that the automatic feature extraction inherent to models based on the DOG paradigm can learn features that are readily interpretable and closely resemble those commonly used in traditional LUR approaches.

  @article{steininger2020maplur, abstract = {Land-use regression (LUR) models are important for the assessment of air pollution concentrations in areas without measurement stations. While many such models exist, they often use manually constructed features based on restricted, locally available data. Thus, they are typically hard to reproduce and challenging to adapt to areas beyond those they have been developed for. In this paper, we advocate a paradigm shift for LUR models: We propose the Data-driven, Open, Global (DOG) paradigm that entails models based on purely data-driven approaches using only openly and globally available data. Progress within this paradigm will alleviate the need for experts to adapt models to the local characteristics of the available data sources and thus facilitate the generalizability of air pollution models to new areas on a global scale. In order to illustrate the feasibility of the DOG paradigm for LUR, we introduce a deep learning model called MapLUR. It is based on a convolutional neural network architecture and is trained exclusively on globally and openly available map data without requiring manual feature engineering. We compare our model to state-of-the-art baselines like linear regression, random forests and multi-layer perceptrons using a large data set of modeled NO2 concentrations in Central London. Our results show that MapLUR significantly outperforms these approaches even though they are provided with manually tailored features. Furthermore, we illustrate that the automatic feature extraction inherent to models based on the DOG paradigm can learn features that are readily interpretable and closely resemble those commonly used in traditional LUR approaches.}, address = {New York, NY, USA}, author = {Steininger, Michael and Kobs, Konstantin and Zehe, Albin and Lautenschlager, Florian and Becker, Martin and Hotho, Andreas}, journal = {ACM Trans. Spatial Algorithms Syst.}, keywords = {regression}, month = {apr}, number = 3, publisher = {Association for Computing Machinery}, title = {MapLUR: Exploring a New Paradigm for Estimating Air Pollution Using Deep Learning on Map Images}, volume = 6, year = 2020 }
• 
  NICER — Aesthetic Image Enhancement with Humans in the Loop M. Fischer; K. Kobs; A. Hotho (2020). 357–362.
  • [ Abstract ]
  • [ BibTeX ]
  Fully- or semi-automatic image enhancement software helps users to increase the visual appeal of photos and does not require in-depth knowledge of manual image editing. However, fully-automatic approaches usually enhance the image in a black-box manner that does not give the user any control over the optimization process, possibly leading to edited images that do not subjectively appeal to the user. Semi-automatic methods mostly allow for controlling which pre-defined editing step is taken, which restricts the users in their creativity and ability to make detailed adjustments, such as brightness or contrast. We argue that incorporating user preferences by guiding an automated enhancement method simplifies image editing and increases the enhancement’s focus on the user. This work thus proposes the Neural Image Correction & Enhancement Routine (NICER), a neural network based approach to no-reference image enhancement in a fully-, semi-automatic or fully manual process that is interactive and user-centered. NICER iteratively adjusts image editing parameters in order to maximize an aesthetic score based on image style and content. Users can modify these parameters at any time and guide the optimization process towards a desired direction. This interactive workflow is a novelty in the field of human-computer interaction for image enhancement tasks. In a user study, we show that NICER can improve image aesthetics without user interaction and that allowing user interaction leads to diverse enhancement outcomes that are strongly preferred over the unedited image. We make our code publicly available to facilitate further research in this direction.

  @inproceedings{fischer2020nicer, abstract = {Fully- or semi-automatic image enhancement software helps users to increase the visual appeal of photos and does not require in-depth knowledge of manual image editing. However, fully-automatic approaches usually enhance the image in a black-box manner that does not give the user any control over the optimization process, possibly leading to edited images that do not subjectively appeal to the user. Semi-automatic methods mostly allow for controlling which pre-defined editing step is taken, which restricts the users in their creativity and ability to make detailed adjustments, such as brightness or contrast. We argue that incorporating user preferences by guiding an automated enhancement method simplifies image editing and increases the enhancement’s focus on the user. This work thus proposes the Neural Image Correction & Enhancement Routine (NICER), a neural network based approach to no-reference image enhancement in a fully-, semi-automatic or fully manual process that is interactive and user-centered. NICER iteratively adjusts image editing parameters in order to maximize an aesthetic score based on image style and content. Users can modify these parameters at any time and guide the optimization process towards a desired direction. This interactive workflow is a novelty in the field of human-computer interaction for image enhancement tasks. In a user study, we show that NICER can improve image aesthetics without user interaction and that allowing user interaction leads to diverse enhancement outcomes that are strongly preferred over the unedited image. We make our code publicly available to facilitate further research in this direction.}, author = {Fischer, Michael and Kobs, Konstantin and Hotho, Andreas}, journal = {ACHI 2020: The Thirteenth International Conference on Advances in Computer-Human Interactions}, keywords = {myown}, pages = {357-362}, title = {NICER — Aesthetic Image Enhancement with Humans in the Loop}, year = 2020 }
• 
  Evaluating the multi-task learning approach for land use regression modelling of air pollution A. Dulny; M. Steininger; F. Lautenschlager; A. Krause; A. Hotho (2020).
  • [ Abstract ]
  • [ BibTeX ]
  Air pollution has been linked to several health problems including heart disease, stroke and lung cancer. Modelling and analyzing this dependency requires reliable and accurate air pollutant measurements collected by stationary air monitoring stations. However, usually only a low number of such stations are present within a single city. To retrieve pollution concentrations for unmeasured locations, researchers rely on land use regression (LUR) models. Those models are typically developed for one pollutant only. However, as results in different areas have shown, modelling several related output variables through multi-task learning can improve the prediction results of the models significantly. In this work, we compared prediction results from single-task and multi-task learning multilayer perceptron models on measurements taken from the OpenSense dataset and the London Atmospheric Emissions Inventory dataset. LUR features were generated from OpenStreetMap using OpenLUR and used to train hard parameter sharing multilayer perceptron models. The results show multi-task learning with sufficient data significantly improves the performance of a LUR model.

  @inproceedings{dulny2020evaluating, abstract = {Air pollution has been linked to several health problems including heart disease, stroke and lung cancer. Modelling and analyzing this dependency requires reliable and accurate air pollutant measurements collected by stationary air monitoring stations. However, usually only a low number of such stations are present within a single city. To retrieve pollution concentrations for unmeasured locations, researchers rely on land use regression (LUR) models. Those models are typically developed for one pollutant only. However, as results in different areas have shown, modelling several related output variables through multi-task learning can improve the prediction results of the models significantly. In this work, we compared prediction results from single-task and multi-task learning multilayer perceptron models on measurements taken from the OpenSense dataset and the London Atmospheric Emissions Inventory dataset. LUR features were generated from OpenStreetMap using OpenLUR and used to train hard parameter sharing multilayer perceptron models. The results show multi-task learning with sufficient data significantly improves the performance of a LUR model.}, author = {Dulny, Andrzej and Steininger, Michael and Lautenschlager, Florian and Krause, Anna and Hotho, Andreas}, booktitle = {International Conference on Frontiers of Artificial Intelligence and Machine Learning}, keywords = {myown}, organization = {IASED}, title = {Evaluating the multi-task learning approach for land use regression modelling of air pollution}, year = 2020 }
• 
  Anomaly Detection in Beehives using Deep Recurrent Autoencoders P. Davidson; M. Steininger; F. Lautenschlager; K. Kobs; A. Krause; A. Hotho (2020). 142–149.
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{davidson2020anomaly, author = {Davidson, Padraig and Steininger, Michael and Lautenschlager, Florian and Kobs, Konstantin and Krause, Anna and Hotho, Andreas}, booktitle = {Proceedings of the 9th International Conference on Sensor Networks (SENSORNETS 2020)}, keywords = {myown}, number = 9, pages = {142-149}, publisher = {SCITEPRESS – Science and Technology Publications, Lda.}, title = {Anomaly Detection in Beehives using Deep Recurrent Autoencoders}, year = 2020 }
• 
  Financial Fraud Detection with Improved Neural Arithmetic Logic Units D. Schlör; M. Ring; A. Krause; A. Hotho (2020). (Vol. Fifth Workshop on MIning DAta for financial applicationS)
  • [ BibTeX ]
  @proceedings{schlor2020financial, author = {Schlör, Daniel and Ring, Markus and Krause, Anna and Hotho, Andreas}, keywords = {myown}, number = 5, title = {Financial Fraud Detection with Improved Neural Arithmetic Logic Units}, volume = {Fifth Workshop on MIning DAta for financial applicationS}, year = 2020 }
• 
  Emote-Controlled: Obtaining Implicit Viewer Feedback through Emote based Sentiment Analysis on Comments of Popular Twitch.tv Channels K. Kobs; A. Zehe; A. Bernstetter; J. Chibane; J. Pfister; J. Tritscher; A. Hotho in {ACM} Transactions on Social Computing (2020). 3(2) 1–34.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  @article{kobs2020emotecontrolled, author = {Kobs, Konstantin and Zehe, Albin and Bernstetter, Armin and Chibane, Julian and Pfister, Jan and Tritscher, Julian and Hotho, Andreas}, journal = {{ACM} Transactions on Social Computing}, keywords = {twitch}, month = {may}, number = 2, pages = {1–34}, publisher = {Association for Computing Machinery ({ACM})}, title = {Emote-Controlled: Obtaining Implicit Viewer Feedback through Emote based Sentiment Analysis on Comments of Popular Twitch.tv Channels}, volume = 3, year = 2020 }
• 
  Smartwatch-Derived Data and Machine Learning Algorithms Estimate Classes of Ratings of Perceived Exertion in Runners: A Pilot Study P. Davidson; P. Düking; C. Zinner; B. Sperlich; A. Hotho in Sensors (2020).
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{davidson2020smartwatchderived, author = {Davidson, Padraig and Düking, Peter and Zinner, Christoph and Sperlich, Billy and Hotho, Andreas}, journal = {Sensors}, keywords = {timeseries}, month = {may}, title = {Smartwatch-Derived Data and Machine Learning Algorithms Estimate Classes of Ratings of Perceived Exertion in Runners: A Pilot Study}, year = 2020 }
• 
  iNALU: Improved Neural Arithmetic Logic Unit D. Schlör; M. Ring; A. Hotho in Frontiers in Artificial Intelligence (2020). 3 71.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Neural networks have to capture mathematical relationships in order to learn various tasks. They approximate these relations implicitly and therefore often do not generalize well. The recently proposed Neural Arithmetic Logic Unit (NALU) is a novel neural architecture which is able to explicitly represent the mathematical relationships by the units of the network to learn operations such as summation, subtraction or multiplication. Although NALUs have been shown to perform well on various downstream tasks, an in-depth analysis reveals practical shortcomings by design, such as the inability to multiply or divide negative input values or training stability issues for deeper networks. We address these issues and propose an improved model architecture. We evaluate our model empirically in various settings from learning basic arithmetic operations to more complex functions. Our experiments indicate that our model solves stability issues and outperforms the original NALU model in means of arithmetic precision and convergence.

  @article{schlor2020inalu, abstract = {Neural networks have to capture mathematical relationships in order to learn various tasks. They approximate these relations implicitly and therefore often do not generalize well. The recently proposed Neural Arithmetic Logic Unit (NALU) is a novel neural architecture which is able to explicitly represent the mathematical relationships by the units of the network to learn operations such as summation, subtraction or multiplication. Although NALUs have been shown to perform well on various downstream tasks, an in-depth analysis reveals practical shortcomings by design, such as the inability to multiply or divide negative input values or training stability issues for deeper networks. We address these issues and propose an improved model architecture. We evaluate our model empirically in various settings from learning basic arithmetic operations to more complex functions. Our experiments indicate that our model solves stability issues and outperforms the original NALU model in means of arithmetic precision and convergence.}, author = {Schlör, Daniel and Ring, Markus and Hotho, Andreas}, journal = {Frontiers in Artificial Intelligence}, keywords = {selected}, pages = 71, title = {iNALU: Improved Neural Arithmetic Logic Unit}, volume = 3, year = 2020 }
• 
  Integrating Keywords into BERT4Rec for Sequential Recommendation E. Fischer; D. Zoller; A. Dallmann; A. Hotho (2020). (Vol. 12325) 275–282.
  • [ Abstract ]
  • [ BibTeX ]
  • [ DOI ]
  • [ Download ]
  A crucial part of recommender systems is to model the user’s preference based on her previous interactions. Different neural networks (e.g., Recurrent Neural Networks), that predict the next item solely based on the sequence of interactions have been successfully applied to sequential recommendation. Recently, BERT4Rec has been proposed, which adapts the BERT architecture based on the Transformer model and training methods used in the Neural Language Modeling community to this task. However, BERT4Rec still only relies on item identifiers to model the user preference, ignoring other sources of information. Therefore, as a first step to include additional information, we propose KeBERT4Rec, a modification of BERT4Rec, which utilizes keyword descriptions of items. We compare two variants for adding keywords to the model on two datasets, a Movielens dataset and a dataset of an online fashion store. First results show that both versions of our model improves the sequential recommending task compared to BERT4Rec.

  @inproceedings{fischer2020integrating, abstract = {A crucial part of recommender systems is to model the user’s preference based on her previous interactions. Different neural networks (e.g., Recurrent Neural Networks), that predict the next item solely based on the sequence of interactions have been successfully applied to sequential recommendation. Recently, BERT4Rec has been proposed, which adapts the BERT architecture based on the Transformer model and training methods used in the Neural Language Modeling community to this task. However, BERT4Rec still only relies on item identifiers to model the user preference, ignoring other sources of information. Therefore, as a first step to include additional information, we propose KeBERT4Rec, a modification of BERT4Rec, which utilizes keyword descriptions of items. We compare two variants for adding keywords to the model on two datasets, a Movielens dataset and a dataset of an online fashion store. First results show that both versions of our model improves the sequential recommending task compared to BERT4Rec.}, address = {Cham}, author = {Fischer, Elisabeth and Zoller, Daniel and Dallmann, Alexander and Hotho, Andreas}, booktitle = {KI 2020: Advances in Artificial Intelligence}, keywords = {sequential}, pages = {275-282}, publisher = {Springer International Publishing}, title = {Integrating Keywords into BERT4Rec for Sequential Recommendation}, volume = 12325, year = 2020 }
• 
  Time Series Forecasting for Self-Aware Systems A. {Bauer}; M. {Züfle}; N. {Herbst}; A. {Zehe}; A. {Hotho}; S. {Kounev} in Proceedings of the IEEE (2020). 1–26.
  • [ Abstract ]
  • [ BibTeX ]
  • [ DOI ]
  Modern distributed systems and Internet-of-Things applications are governed by fast living and changing requirements. Moreover, they have to struggle with huge amounts of data that they create or have to process. To improve the self-awareness of such systems and enable proactive and autonomous decisions, reliable time series forecasting methods are required. However, selecting a suitable forecasting method for a given scenario is a challenging task. According to the ``No-Free-Lunch Theorem,'' there is no general forecasting method that always performs best. Thus, manual feature engineering remains to be a mandatory expert task to avoid trial and error. Furthermore, determining the expected time-to-result of existing forecasting methods is a challenge. In this article, we extensively assess the state-of-the-art in time series forecasting. We compare existing methods and discuss the issues that have to be addressed to enable their use in a self-aware computing context. To address these issues, we present a step-by-step approach to fully automate the feature engineering and forecasting process. Then, following the principles from benchmarking, we establish a level-playing field for evaluating the accuracy and time-to-result of automated forecasting methods for a broad set of application scenarios. We provide results of a benchmarking competition to guide in selecting and appropriately using existing forecasting methods for a given self-aware computing context. Finally, we present a case study in the area of self-aware data-center resource management to exemplify the benefits of fully automated learning and reasoning processes on time series data.

  @article{bauer2020series, abstract = {Modern distributed systems and Internet-of-Things applications are governed by fast living and changing requirements. Moreover, they have to struggle with huge amounts of data that they create or have to process. To improve the self-awareness of such systems and enable proactive and autonomous decisions, reliable time series forecasting methods are required. However, selecting a suitable forecasting method for a given scenario is a challenging task. According to the ``No-Free-Lunch Theorem,'' there is no general forecasting method that always performs best. Thus, manual feature engineering remains to be a mandatory expert task to avoid trial and error. Furthermore, determining the expected time-to-result of existing forecasting methods is a challenge. In this article, we extensively assess the state-of-the-art in time series forecasting. We compare existing methods and discuss the issues that have to be addressed to enable their use in a self-aware computing context. To address these issues, we present a step-by-step approach to fully automate the feature engineering and forecasting process. Then, following the principles from benchmarking, we establish a level-playing field for evaluating the accuracy and time-to-result of automated forecasting methods for a broad set of application scenarios. We provide results of a benchmarking competition to guide in selecting and appropriately using existing forecasting methods for a given self-aware computing context. Finally, we present a case study in the area of self-aware data-center resource management to exemplify the benefits of fully automated learning and reasoning processes on time series data.}, author = {{Bauer}, A. and {Züfle}, M. and {Herbst}, N. and {Zehe}, A. and {Hotho}, A. and {Kounev}, S.}, journal = {Proceedings of the IEEE}, keywords = {timeseries}, pages = {1-26}, title = {Time Series Forecasting for Self-Aware Systems}, year = 2020 }
• 
  Improving Sentiment Analysis with Biofeedback Data D. Schl{\"o}r; A. Zehe; K. Kobs; B. Veseli; F. Westermeier; L. Br{\"u}bach; D. Roth; M. E. Latoschik; A. Hotho (2020). 28–33.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Humans frequently are able to read and interpret emotions of others by directly taking verbal and non-verbal signals in human-to-human communication into account or to infer or even experience emotions from mediated stories. For computers, however, emotion recognition is a complex problem: Thoughts and feelings are the roots of many behavioural responses and they are deeply entangled with neurophysiological changes within humans. As such, emotions are very subjective, often are expressed in a subtle manner, and are highly depending on context. For example, machine learning approaches for text-based sentiment analysis often rely on incorporating sentiment lexicons or language models to capture the contextual meaning. This paper explores if and how we further can enhance sentiment analysis using biofeedback of humans which are experiencing emotions while reading texts. Specifically, we record the heart rate and brain waves of readers that are presented with short texts which have been annotated with the emotions they induce. We use these physiological signals to improve the performance of a lexicon-based sentiment classifier. We find that the combination of several biosignals can improve the ability of a text-based classifier to detect the presence of a sentiment in a text on a per-sentence level.

  @inproceedings{schlor2020improving, abstract = {Humans frequently are able to read and interpret emotions of others by directly taking verbal and non-verbal signals in human-to-human communication into account or to infer or even experience emotions from mediated stories. For computers, however, emotion recognition is a complex problem: Thoughts and feelings are the roots of many behavioural responses and they are deeply entangled with neurophysiological changes within humans. As such, emotions are very subjective, often are expressed in a subtle manner, and are highly depending on context. For example, machine learning approaches for text-based sentiment analysis often rely on incorporating sentiment lexicons or language models to capture the contextual meaning. This paper explores if and how we further can enhance sentiment analysis using biofeedback of humans which are experiencing emotions while reading texts. Specifically, we record the heart rate and brain waves of readers that are presented with short texts which have been annotated with the emotions they induce. We use these physiological signals to improve the performance of a lexicon-based sentiment classifier. We find that the combination of several biosignals can improve the ability of a text-based classifier to detect the presence of a sentiment in a text on a per-sentence level.}, address = {Marseille, France}, author = {Schl{\"o}r, Daniel and Zehe, Albin and Kobs, Konstantin and Veseli, Blerta and Westermeier, Franziska and Br{\"u}bach, Larissa and Roth, Daniel and Latoschik, Marc Erich and Hotho, Andreas}, booktitle = {Proceedings of LREC2020 Workshop ``People in language, vision and the mind'' (ONION2020)}, keywords = {sentimentanalysis}, month = {may}, pages = {28–33}, publisher = {European Language Resources Association (ELRA)}, title = {Improving Sentiment Analysis with Biofeedback Data}, year = 2020 }
• 
  HarryMotions – Classifying Relationships in Harry Potter based on Emotion Analysis A. Zehe; J. Arns; L. Hettinger; A. Hotho (2020).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{zehe2020harrymotions, author = {Zehe, Albin and Arns, Julia and Hettinger, Lena and Hotho, Andreas}, booktitle = {5th SwissText & 16th KONVENS Joint Conference}, keywords = {sentimentanalysis}, title = {HarryMotions – Classifying Relationships in Harry Potter based on Emotion Analysis}, year = 2020 }
• 
  Towards Predicting the Subscription Status of Twitch.tv Users K. Kobs; M. Potthast; M. Wiegmann; A. Zehe; B. Stein; A. Hotho in Proceedings of ECML-PKDD 2020 ChAT Discovery Challenge on Chat Analytics for Twitch (2020).
  • [ BibTeX ]
  • [ URL ]
  @article{kobs2020towards, author = {Kobs, Konstantin and Potthast, Martin and Wiegmann, Matti and Zehe, Albin and Stein, Benno and Hotho, Andreas}, journal = {Proceedings of ECML-PKDD 2020 ChAT Discovery Challenge on Chat Analytics for Twitch}, keywords = {twitch}, title = {Towards Predicting the Subscription Status of Twitch.tv Users}, year = 2020 }

• 
  EClaiRE: Context Matters! – Comparing Word Embeddings for Relation Classification L. Hettinger; A. Zehe; A. Dallmann; A. Hotho K. David, K. Geihs, M. Lange, G. Stumme (Eds.) (2019). 191–204.
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{hettinger2019eclaire, address = {Bonn}, author = {Hettinger, Lena and Zehe, Albin and Dallmann, Alexander and Hotho, Andreas}, booktitle = {INFORMATIK 2019: 50 Jahre Gesellschaft für Informatik – Informatik für Gesellschaft}, editor = {David, Klaus and Geihs, Kurt and Lange, Martin and Stumme, Gerd}, keywords = {word_embedding}, pages = {191-204}, publisher = {Gesellschaft für Informatik e.V.}, title = {EClaiRE: Context Matters! – Comparing Word Embeddings for Relation Classification}, year = 2019 }
• 
  Flow-based network traffic generation using Generative Adversarial Networks M. Ring; D. Schlör; D. Landes; A. Hotho in Computers & Security (2019). 82 156–172.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Flow-based data sets are necessary for evaluating network-based intrusion detection systems (NIDS). In this work, we propose a novel methodology for generating realistic flow-based network traffic. Our approach is based on Generative Adversarial Networks (GANs) which achieve good results for image generation. A major challenge lies in the fact that GANs can only process continuous attributes. However, flow-based data inevitably contain categorical attributes such as IP addresses or port numbers. Therefore, we propose three different preprocessing approaches for flow-based data in order to transform them into continuous values. Further, we present a new method for evaluating the generated flow-based network traffic which uses domain knowledge to define quality tests. We use the three approaches for generating flow-based network traffic based on the CIDDS-001 data set. Experiments indicate that two of the three approaches are able to generate high quality data.

  @article{RING2019156, abstract = {Flow-based data sets are necessary for evaluating network-based intrusion detection systems (NIDS). In this work, we propose a novel methodology for generating realistic flow-based network traffic. Our approach is based on Generative Adversarial Networks (GANs) which achieve good results for image generation. A major challenge lies in the fact that GANs can only process continuous attributes. However, flow-based data inevitably contain categorical attributes such as IP addresses or port numbers. Therefore, we propose three different preprocessing approaches for flow-based data in order to transform them into continuous values. Further, we present a new method for evaluating the generated flow-based network traffic which uses domain knowledge to define quality tests. We use the three approaches for generating flow-based network traffic based on the CIDDS-001 data set. Experiments indicate that two of the three approaches are able to generate high quality data.}, author = {Ring, Markus and Schlör, Daniel and Landes, Dieter and Hotho, Andreas}, journal = {Computers & Security}, keywords = {traffic}, pages = {156 - 172}, title = {Flow-based network traffic generation using Generative Adversarial Networks}, volume = 82, year = 2019 }
• 
  On the Right Track! Analysing and Predicting Navigation Success in Wikipedia T. Koopmann; A. Dallmann; L. Hettinger; T. Niebler; A. Hotho in HT ’19, C. Atzenbeck, J. Rubart, D. E. Millard (Eds.) (2019). 143–152.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  @inproceedings{Koopmann:2019:RTA:3342220.3343650, address = {New York, NY, USA}, author = {Koopmann, Tobias and Dallmann, Alexander and Hettinger, Lena and Niebler, Thomas and Hotho, Andreas}, booktitle = {Proceedings of the 30th ACM Conference on Hypertext and Social Media}, editor = {Atzenbeck, Claus and Rubart, Jessica and Millard, David E.}, keywords = {wikipedia}, pages = {143–152}, publisher = {ACM}, series = {HT '19}, title = {On the Right Track! Analysing and Predicting Navigation Success in Wikipedia}, year = 2019 }
• 
  Comparison of System Call Representations for Intrusion Detection. S. Wunderlich; M. Ring; D. Landes; A. Hotho in Advances in Intelligent Systems and Computing, F. Martínez-Álvarez, A. T. Lora, J. A. S. Muñoz, H. Quintián, E. Corchado (Eds.) (2019). (Vol. 951) 14–24.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{wunderlich2019comparison, author = {Wunderlich, Sarah and Ring, Markus and Landes, Dieter and Hotho, Andreas}, booktitle = {CISIS-ICEUTE}, crossref = {conf/cisis-spain/2019}, editor = {Martínez-Álvarez, Francisco and Lora, Alicia Troncoso and Muñoz, José António Sáez and Quintián, Héctor and Corchado, Emilio}, keywords = {system}, pages = {14-24}, publisher = {Springer}, series = {Advances in Intelligent Systems and Computing}, title = {Comparison of System Call Representations for Intrusion Detection.}, volume = 951, year = 2019 }
• 
  Team Xenophilius Lovegood at SemEval-2019 Task 4: Hyperpartisanship Classification using Convolutional Neural Networks A. Zehe; L. Hettinger; S. Ernst; C. Hauptmann; A. Hotho (2019).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{zehe2019xenophilius, author = {Zehe, Albin and Hettinger, Lena and Ernst, Stefan and Hauptmann, Christian and Hotho, Andreas}, booktitle = {Proceedings of The 13th International Workshop on Semantic Evaluation}, keywords = {semeval}, publisher = {Association for Computational Linguistics}, title = {Team Xenophilius Lovegood at SemEval-2019 Task 4: Hyperpartisanship Classification using Convolutional Neural Networks}, year = 2019 }
• 
  Solving Mathematical Exercises: Prediction of Students’ Success S. Wankerl; G. Götz; A. Hotho in CEUR Workshop Proceedings, R. Jäschke, M. Weidlich (Eds.) (2019). (Vol. 2454) 190–194.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{wankerl2019solving, author = {Wankerl, Sebastian and Götz, Gerhard and Hotho, Andreas}, booktitle = {LWDA}, editor = {Jäschke, Robert and Weidlich, Matthias}, keywords = {myown}, pages = {190-194}, publisher = {CEUR-WS.org}, series = {CEUR Workshop Proceedings}, title = {Solving Mathematical Exercises: Prediction of Students’ Success}, volume = 2454, year = 2019 }
• 
  A Survey of Network-based Intrusion Detection Data Sets M. Ring; S. Wunderlich; D. Scheuring; D. Landes; A. Hotho in Computers & Security (2019). 86 147–167.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{ring2019survey, author = {Ring, Markus and Wunderlich, Sarah and Scheuring, Deniz and Landes, Dieter and Hotho, Andreas}, journal = {Computers & Security}, keywords = {myown}, pages = {147-167}, publisher = {Elsevier Advanced Technology}, title = {A Survey of Network-based Intrusion Detection Data Sets}, volume = 86, year = 2019 }
• 
  Detection of Scenes in Fiction E. Gius; F. Jannidis; M. Krug; A. Zehe; A. Hotho; F. Puppe; J. Krebs; N. Reiter; N. Wiedmer; L. Konle (2019).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{gius2019detection, author = {Gius, Evelyn and Jannidis, Fotis and Krug, Markus and Zehe, Albin and Hotho, Andreas and Puppe, Frank and Krebs, Jonathan and Reiter, Nils and Wiedmer, Nathalie and Konle, Leonard}, booktitle = {Proceedings of Digital Humanities 2019}, keywords = {scene}, title = {Detection of Scenes in Fiction}, year = 2019 }

• 
  Flow-based Network Traffic Generation using Generative Adversarial Networks. M. Ring; D. Schlör; D. Landes; A. Hotho in CoRR (2018). abs/1810.07795
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Flow-based data sets are necessary for evaluating network-based intrusion de- tection systems (NIDS). In this work, we propose a novel methodology for gener- ating realistic flow-based network traffic. Our approach is based on Generative Adversarial Networks (GANs) which achieve good results for image generation. A major challenge lies in the fact that GANs can only process continuous at- tributes. However, flow-based data inevitably contain categorical attributes such as IP addresses or port numbers. Therefore, we propose three different preprocessing approaches for flow-based data in order to transform them into continuous values. Further, we present a new method for evaluating the gener- ated flow-based network traffic which uses domain knowledge to define quality tests. We use the three approaches for generating flow-based network traffic based on the CIDDS-001 data set. Experiments indicate that two of the three approaches are able to generate high quality data.

  @article{journals/corr/abs-1810-07795, abstract = {Flow-based data sets are necessary for evaluating network-based intrusion de- tection systems (NIDS). In this work, we propose a novel methodology for gener- ating realistic flow-based network traffic. Our approach is based on Generative Adversarial Networks (GANs) which achieve good results for image generation. A major challenge lies in the fact that GANs can only process continuous at- tributes. However, flow-based data inevitably contain categorical attributes such as IP addresses or port numbers. Therefore, we propose three different preprocessing approaches for flow-based data in order to transform them into continuous values. Further, we present a new method for evaluating the gener- ated flow-based network traffic which uses domain knowledge to define quality tests. We use the three approaches for generating flow-based network traffic based on the CIDDS-001 data set. Experiments indicate that two of the three approaches are able to generate high quality data.}, author = {Ring, Markus and Schlör, Daniel and Landes, Dieter and Hotho, Andreas}, journal = {CoRR}, keywords = {traffic}, title = {Flow-based Network Traffic Generation using Generative Adversarial Networks.}, volume = {abs/1810.07795}, year = 2018 }
• 
  ClaiRE at SemEval-2018 Task 7: Classification of Relations using Embeddings L. Hettinger; A. Dallmann; A. Zehe; T. Niebler; A. Hotho (2018).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{hettinger2018semeval, address = {New Orleans, LA, USA}, author = {Hettinger, Lena and Dallmann, Alexander and Zehe, Albin and Niebler, Thomas and Hotho, Andreas}, booktitle = {Proceedings of International Workshop on Semantic Evaluation (SemEval-2018)}, keywords = {word}, title = {ClaiRE at SemEval-2018 Task 7: Classification of Relations using Embeddings}, year = 2018 }
• 
  {Adaptive kNN Using Expected Accuracy for Classification of Geo-Spatial Data} M. Kibanov; M. Becker; J. M\"uller; M. Atzmueller; A. Hotho; G. Stumme (2018).
  • [ BibTeX ]
  @inproceedings{KBMAHS:18, address = {New York, NY, USA}, author = {Kibanov, Mark and Becker, Martin and M\"uller, J\"urgen and Atzmueller, Martin and Hotho, Andreas and Stumme, Gerd}, booktitle = {Proc. 33rd ACM Symposium On Applied Computing}, keywords = {spatial}, publisher = {ACM Press}, title = {{Adaptive kNN Using Expected Accuracy for Classification of Geo-Spatial Data}}, year = 2018 }
• 
  Accessing Information with Tags: Search and Ranking B. Navarro Bullock; A. Hotho; G. Stumme in Social Information Access: Systems and Technologies, P. Brusilovsky, D. He (Eds.) (2018). 310–343.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  With the growth of the Social Web, a variety of new web-based services arose and changed the way users interact with the internet and consume information. One central phenomenon was and is tagging which allows to manage, organize and access information in social systems. Tagging helps to manage all kinds of resources, making their access much easier. The first type of social tagging systems were social bookmarking systems, i.e., platforms for storing and sharing bookmarks on the web rather than just in the browser. Meanwhile, (hash-)tagging is central in many other Social Media systems such as social networking sites and micro-blogging platforms. To allow for efficient information access, special algorithms have been developed to guide the user, to search for information and to rank the content based on tagging information contributed by the users.

  @inbook{NavarroBullock2018, abstract = {With the growth of the Social Web, a variety of new web-based services arose and changed the way users interact with the internet and consume information. One central phenomenon was and is tagging which allows to manage, organize and access information in social systems. Tagging helps to manage all kinds of resources, making their access much easier. The first type of social tagging systems were social bookmarking systems, i.e., platforms for storing and sharing bookmarks on the web rather than just in the browser. Meanwhile, (hash-)tagging is central in many other Social Media systems such as social networking sites and micro-blogging platforms. To allow for efficient information access, special algorithms have been developed to guide the user, to search for information and to rank the content based on tagging information contributed by the users.}, address = {Cham}, author = {Navarro Bullock, Beate and Hotho, Andreas and Stumme, Gerd}, booktitle = {Social Information Access: Systems and Technologies}, editor = {Brusilovsky, Peter and He, Daqing}, keywords = {tag}, pages = {310–343}, publisher = {Springer International Publishing}, title = {Accessing Information with Tags: Search and Ranking}, year = 2018 }
• 
  Big-Data Helps SDN to Improve Application Specific Quality of Service S. Schwarzmann; A. Blenk; O. Dobrijevic; M. Jarschel; A. Hotho; T. Zinner; F. Wamser in Big Data and Software Defined Networks (2018).
  • [ Abstract ]
  • [ BibTeX ]
  This chapter first provides an outline of the current results in the domains of: (a) quality-of-service (QoS) / quality-of-experience (QoE) control and management (CaM) for real-time multimedia services that is supported by software-defined networking (SDN), and (b) big data analytics and methods that are used for QoS/QoE CaM. Then, three specific use case scenarios with respect to video streaming services are presented, so as to illustrate the expected benefits of incorporating big data analytics into SDN-based CaM for the purposes of improving or optimizing QoS/QoE. In the end, we describe our vision and a high-level view of an SDN-based architecture for QoS/QoE CaM that is enriched with big data analytics' functional blocks and summarize corresponding challenges.

  @incollection{info3-incollection-2018-2, abstract = {This chapter first provides an outline of the current results in the domains of: (a) quality-of-service (QoS) / quality-of-experience (QoE) control and management (CaM) for real-time multimedia services that is supported by software-defined networking (SDN), and (b) big data analytics and methods that are used for QoS/QoE CaM. Then, three specific use case scenarios with respect to video streaming services are presented, so as to illustrate the expected benefits of incorporating big data analytics into SDN-based CaM for the purposes of improving or optimizing QoS/QoE. In the end, we describe our vision and a high-level view of an SDN-based architecture for QoS/QoE CaM that is enriched with big data analytics' functional blocks and summarize corresponding challenges.}, author = {Schwarzmann, Susanna and Blenk, Andreas and Dobrijevic, Ognjen and Jarschel, Michael and Hotho, Andreas and Zinner, Thomas and Wamser, Florian}, booktitle = {Big Data and Software Defined Networks}, keywords = {service}, publisher = {IET}, title = {Big-Data Helps SDN to Improve Application Specific Quality of Service}, year = 2018 }
• 
  Burrows Zeta: Varianten und Evaluation C. Schöch; J. Calvo; A. Zehe; A. Hotho (2018).
  • [ BibTeX ]
  @inproceedings{schoch2018burrows, author = {Schöch, Christof and Calvo, José and Zehe, Albin and Hotho, Andreas}, booktitle = {DHd 2018}, keywords = {zeta}, title = {Burrows Zeta: Varianten und Evaluation}, year = 2018 }
• 
  Detection of slow port scans in flow-based network traffic M. Ring; D. Landes; A. Hotho in PLOS ONE (2018). 13(9) 1–18.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Frequently, port scans are early indicators of more serious attacks. Unfortunately, the detection of slow port scans in company networks is challenging due to the massive amount of network data. This paper proposes an innovative approach for preprocessing flow-based data which is specifically tailored to the detection of slow port scans. The preprocessing chain generates new objects based on flow-based data aggregated over time windows while taking domain knowledge as well as additional knowledge about the network structure into account. The computed objects are used as input for the further analysis. Based on these objects, we propose two different approaches for detection of slow port scans. One approach is unsupervised and uses sequential hypothesis testing whereas the other approach is supervised and uses classification algorithms. We compare both approaches with existing port scan detection algorithms on the flow-based CIDDS-001 data set. Experiments indicate that the proposed approaches achieve better detection rates and exhibit less false alarms than similar algorithms.

  @article{10.1371/journal.pone.0204507, abstract = {Frequently, port scans are early indicators of more serious attacks. Unfortunately, the detection of slow port scans in company networks is challenging due to the massive amount of network data. This paper proposes an innovative approach for preprocessing flow-based data which is specifically tailored to the detection of slow port scans. The preprocessing chain generates new objects based on flow-based data aggregated over time windows while taking domain knowledge as well as additional knowledge about the network structure into account. The computed objects are used as input for the further analysis. Based on these objects, we propose two different approaches for detection of slow port scans. One approach is unsupervised and uses sequential hypothesis testing whereas the other approach is supervised and uses classification algorithms. We compare both approaches with existing port scan detection algorithms on the flow-based CIDDS-001 data set. Experiments indicate that the proposed approaches achieve better detection rates and exhibit less false alarms than similar algorithms.}, author = {Ring, Markus and Landes, Dieter and Hotho, Andreas}, journal = {PLOS ONE}, keywords = {traffic}, month = {09}, number = 9, pages = {1-18}, publisher = {Public Library of Science}, title = {Detection of slow port scans in flow-based network traffic}, volume = 13, year = 2018 }
• 
  EveryAware Gears: A Tool to visualize and analyze all types of Citizen Science Data F. Lautenschlager; M. Becker; M. Steininger; A. Hotho D. Burghardt, S. Chen, G. Andrienko, N. Andrienko, R. Purves, A. Diehl (Eds.) (2018).
  • [ BibTeX ]
  @inproceedings{lautenschlager2018everyaware, author = {Lautenschlager, Florian and Becker, Martin and Steininger, Michael and Hotho, Andreas}, booktitle = {Proceedings of VGI Geovisual Analytics Workshop, colocated with BDVA 2018}, editor = {Burghardt, Dirk and Chen, Siming and Andrienko, Gennady and Andrienko, Nathalia and Purves, Ross and Diehl, Alexandra}, keywords = {tool}, month = {oct}, publisher = {KOPS}, title = {EveryAware Gears: A Tool to visualize and analyze all types of Citizen Science Data}, year = 2018 }
• 
  Air Trails -- Urban Air Quality Campaign Exploration Patterns M. Becker; F. Lautenschlager; A. Hotho (2018).
  • [ BibTeX ]
  @article{becker2018air, author = {Becker, Martin and Lautenschlager, Florian and Hotho, Andreas}, keywords = {project:p2map}, title = {Air Trails – Urban Air Quality Campaign Exploration Patterns}, year = 2018 }
• 
  Burrows’ Zeta: Exploring and Evaluating Variants and Parameters C. Schöch; D. Schlör; A. Zehe; H. Gebhard; M. Becker; A. Hotho (2018). 274–277.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{schoech2018zeta, author = {Schöch, Christof and Schlör, Daniel and Zehe, Albin and Gebhard, Henning and Becker, Martin and Hotho, Andreas}, booktitle = {DH}, keywords = {zeta}, pages = {274-277}, title = {Burrows’ Zeta: Exploring and Evaluating Variants and Parameters}, year = 2018 }
• 
  A White-Box Model for Detecting Author Nationality by Linguistic Differences in Spanish Novels A. Zehe; D. Schlör; U. Henny-Krahmer; M. Becker; A. Hotho (2018).
  • [ BibTeX ]
  @inproceedings{zehe2018whitebox, author = {Zehe, Albin and Schlör, Daniel and Henny-Krahmer, Ulrike and Becker, Martin and Hotho, Andreas}, booktitle = {DH}, keywords = {nationality}, organization = {ADHO}, title = {A White-Box Model for Detecting Author Nationality by Linguistic Differences in Spanish Novels}, year = 2018 }
• 
  Analysing Direct Speech in German Novels F. Jannidis; L. Konle; A. Zehe; A. Hotho; M. Krug (2018).
  • [ BibTeX ]
  @inproceedings{jannidis2018analysing, author = {Jannidis, Fotis and Konle, Leonard and Zehe, Albin and Hotho, Andreas and Krug, Markus}, booktitle = {DHd 2018}, keywords = {myown}, title = {Analysing Direct Speech in German Novels}, year = 2018 }

• 
  Participatory sensing, opinions and collective awareness V. Loreto; M. Haklay; A. Hotho; V. C. P. Servedio; G. Stumme; J. Theunis; F. Tria (2017). Springer.
  • [ BibTeX ]
  • [ Download ]
  @book{loreto2017participatory, editor = {Loreto, Vittorio and Haklay, Mordechai and Hotho, Andreas and Servedio, Vito C. P. and Stumme, Gerd and Theunis, Jan and Tria, Francesca}, keywords = {sensing}, publisher = {Springer}, title = {Participatory sensing, opinions and collective awareness}, year = 2017 }
• 
  Learning Semantic Relatedness From Human Feedback Using Metric Learning T. Niebler; M. Becker; C. Pölitz; A. Hotho (2017).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Assessing the degree of semantic relatedness between words is an important task with a variety of semantic applications, such as ontology learning for the Semantic Web, semantic search or query expansion. To accomplish this in an automated fashion, many relatedness measures have been proposed. However, most of these metrics only encode information contained in the underlying corpus and thus do not directly model human intuition. To solve this, we propose to utilize a metric learning approach to improve existing semantic relatedness measures by learning from additional information, such as explicit human feedback. For this, we argue to use word embeddings instead of traditional high-dimensional vector representations in order to leverage their semantic density and to reduce computational cost. We rigorously test our approach on several domains including tagging data as well as publicly available embeddings based on Wikipedia texts and navigation. Human feedback about semantic relatedness for learning and evaluation is extracted from publicly available datasets such as MEN or WS-353. We find that our method can significantly improve semantic relatedness measures by learning from additional information, such as explicit human feedback. For tagging data, we are the first to generate and study embeddings. Our results are of special interest for ontology and recommendation engineers, but also for any other researchers and practitioners of Semantic Web techniques.

  @misc{niebler2017learning, abstract = {Assessing the degree of semantic relatedness between words is an important task with a variety of semantic applications, such as ontology learning for the Semantic Web, semantic search or query expansion. To accomplish this in an automated fashion, many relatedness measures have been proposed. However, most of these metrics only encode information contained in the underlying corpus and thus do not directly model human intuition. To solve this, we propose to utilize a metric learning approach to improve existing semantic relatedness measures by learning from additional information, such as explicit human feedback. For this, we argue to use word embeddings instead of traditional high-dimensional vector representations in order to leverage their semantic density and to reduce computational cost. We rigorously test our approach on several domains including tagging data as well as publicly available embeddings based on Wikipedia texts and navigation. Human feedback about semantic relatedness for learning and evaluation is extracted from publicly available datasets such as MEN or WS-353. We find that our method can significantly improve semantic relatedness measures by learning from additional information, such as explicit human feedback. For tagging data, we are the first to generate and study embeddings. Our results are of special interest for ontology and recommendation engineers, but also for any other researchers and practitioners of Semantic Web techniques.}, author = {Niebler, Thomas and Becker, Martin and Pölitz, Christian and Hotho, Andreas}, keywords = {semantic}, note = {cite arxiv:1705.07425Comment: Under review at ISWC 2017}, title = {Learning Semantic Relatedness From Human Feedback Using Metric Learning}, year = 2017 }
• 
  Mining social semantics on the social web A. Hotho; R. Jaeschke; K. Lerman in Semantic Web (2017). 8(5) 623–624.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @article{wrro115871, author = {Hotho, A. and Jaeschke, R. and Lerman, K.}, journal = {Semantic Web}, keywords = {web}, month = {April}, note = {\copyright 2017 IOS Press and the authors. This is an author produced version of a paper subsequently published in Semantic Web. Uploaded in accordance with the publisher's self-archiving policy.}, number = 5, pages = {623–624}, publisher = {IOS Press}, title = {Mining social semantics on the social web}, volume = 8, year = 2017 }
• 
  Learning Semantic Relatedness from Human Feedback Using Relative Relatedness Learning T. Niebler; M. Becker; C. Pölitz; A. Hotho (2017).
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{niebler2017learning, author = {Niebler, Thomas and Becker, Martin and Pölitz, Christian and Hotho, Andreas}, booktitle = {Proceedings of the ISWC 2017 Posters & Demonstrations and Industry Tracks co-located with 16th International Semantic Web Conference (ISWC 2017)}, keywords = {semantics}, title = {Learning Semantic Relatedness from Human Feedback Using Relative Relatedness Learning}, year = 2017 }
• 
  MixedTrails: Bayesian hypothesis comparison on heterogeneous sequential data M. Becker; F. Lemmerich; P. Singer; M. Strohmaier; A. Hotho in Data Mining and Knowledge Discovery (2017).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Sequential traces of user data are frequently observed online and offline, e.g., as sequences of visited websites or as sequences of locations captured by GPS. However, understanding factors explaining the production of sequence data is a challenging task, especially since the data generation is often not homogeneous. For example, navigation behavior might change in different phases of browsing a website or movement behavior may vary between groups of users. In this work, we tackle this task and propose MixedTrails , a Bayesian approach for comparing the plausibility of hypotheses regarding the generative processes of heterogeneous sequence data. Each hypothesis is derived from existing literature, theory, or intuition and represents a belief about transition probabilities between a set of states that can vary between groups of observed transitions. For example, when trying to understand human movement in a city and given some data, a hypothesis assuming tourists to be more likely to move towards points of interests than locals can be shown to be more plausible than a hypothesis assuming the opposite. Our approach incorporates such hypotheses as Bayesian priors in a generative mixed transition Markov chain model, and compares their plausibility utilizing Bayes factors. We discuss analytical and approximate inference methods for calculating the marginal likelihoods for Bayes factors, give guidance on interpreting the results, and illustrate our approach with several experiments on synthetic and empirical data from Wikipedia and Flickr. Thus, this work enables a novel kind of analysis for studying sequential data in many application areas.

  @article{Becker2017, abstract = {Sequential traces of user data are frequently observed online and offline, e.g., as sequences of visited websites or as sequences of locations captured by GPS. However, understanding factors explaining the production of sequence data is a challenging task, especially since the data generation is often not homogeneous. For example, navigation behavior might change in different phases of browsing a website or movement behavior may vary between groups of users. In this work, we tackle this task and propose MixedTrails , a Bayesian approach for comparing the plausibility of hypotheses regarding the generative processes of heterogeneous sequence data. Each hypothesis is derived from existing literature, theory, or intuition and represents a belief about transition probabilities between a set of states that can vary between groups of observed transitions. For example, when trying to understand human movement in a city and given some data, a hypothesis assuming tourists to be more likely to move towards points of interests than locals can be shown to be more plausible than a hypothesis assuming the opposite. Our approach incorporates such hypotheses as Bayesian priors in a generative mixed transition Markov chain model, and compares their plausibility utilizing Bayes factors. We discuss analytical and approximate inference methods for calculating the marginal likelihoods for Bayes factors, give guidance on interpreting the results, and illustrate our approach with several experiments on synthetic and empirical data from Wikipedia and Flickr. Thus, this work enables a novel kind of analysis for studying sequential data in many application areas.}, author = {Becker, Martin and Lemmerich, Florian and Singer, Philipp and Strohmaier, Markus and Hotho, Andreas}, journal = {Data Mining and Knowledge Discovery}, keywords = {myown}, month = {jul}, title = {MixedTrails: Bayesian hypothesis comparison on heterogeneous sequential data}, year = 2017 }
• 
  Leveraging User-Interactions for Time-Aware Tag Recommendations D. Zoller; S. Doerfel; C. Pölitz; A. Hotho in {CEUR} Workshop Proceedings (2017).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{zoller2017leveraging, author = {Zoller, Daniel and Doerfel, Stephan and Pölitz, Christian and Hotho, Andreas}, booktitle = {Proceedings of the Workshop on Temporal Reasoning in Recommender Systems}, keywords = {user}, series = {{CEUR} Workshop Proceedings}, title = {Leveraging User-Interactions for Time-Aware Tag Recommendations}, year = 2017 }
• 
  A Toolset for Intrusion and Insider Threat Detection M. Ring; S. Wunderlich; D. Gr{\"u}dl; D. Landes; A. Hotho in Data Analytics and Decision Support for Cybersecurity: Trends, Methodologies and Applications, I. Palomares Carrascosa, H. K. Kalutarage, Y. Huang (Eds.) (2017). 3–31.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Company data are a valuable asset and must be protected against unauthorized access and manipulation. In this contribution, we report on our ongoing work that aims to support IT security experts with identifying novel or obfuscated attacks in company networks, irrespective of their origin inside or outside the company network. A new toolset for anomaly based network intrusion detection is proposed. This toolset uses flow-based data which can be easily retrieved by central network components. We study the challenges of analysing flow-based data streams using data mining algorithms and build an appropriate approach step by step. In contrast to previous work, we collect flow-based data for each host over a certain time window, include the knowledge of domain experts and analyse the data from three different views. We argue that incorporating expert knowledge and previous flows allow us to create more meaningful attributes for subsequent analysis methods. This way, we try to detect novel attacks while simultaneously limiting the number of false positives.

  @inbook{Ring2017, abstract = {Company data are a valuable asset and must be protected against unauthorized access and manipulation. In this contribution, we report on our ongoing work that aims to support IT security experts with identifying novel or obfuscated attacks in company networks, irrespective of their origin inside or outside the company network. A new toolset for anomaly based network intrusion detection is proposed. This toolset uses flow-based data which can be easily retrieved by central network components. We study the challenges of analysing flow-based data streams using data mining algorithms and build an appropriate approach step by step. In contrast to previous work, we collect flow-based data for each host over a certain time window, include the knowledge of domain experts and analyse the data from three different views. We argue that incorporating expert knowledge and previous flows allow us to create more meaningful attributes for subsequent analysis methods. This way, we try to detect novel attacks while simultaneously limiting the number of false positives.}, address = {Cham}, author = {Ring, Markus and Wunderlich, Sarah and Gr{\"u}dl, Dominik and Landes, Dieter and Hotho, Andreas}, booktitle = {Data Analytics and Decision Support for Cybersecurity: Trends, Methodologies and Applications}, editor = {Palomares Carrascosa, Iv{\'a}n and Kalutarage, Harsha Kumara and Huang, Yan}, keywords = {toolset}, pages = {3–31}, publisher = {Springer International Publishing}, title = {A Toolset for Intrusion and Insider Threat Detection}, year = 2017 }
• 
  Learning Word Embeddings from Tagging Data: A methodological comparison T. Niebler; L. Hahn; A. Hotho (2017).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{niebler2017learning_2, author = {Niebler, Thomas and Hahn, Luzian and Hotho, Andreas}, booktitle = {Proceedings of the LWDA}, keywords = {word2vec}, title = {Learning Word Embeddings from Tagging Data: A methodological comparison}, year = 2017 }
• 
  A Bayesian Method for Comparing Hypotheses About Human Trails P. Singer; D. Helic; A. Hotho; M. Strohmaier in ACM Trans. Web (2017). 11(3) 14:1–14:29.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  When users interact with the Web today, they leave sequential digital trails on a massive scale. Examples of such human trails include Web navigation, sequences of online restaurant reviews, or online music play lists. Understanding the factors that drive the production of these trails can be useful, for example, for improving underlying network structures, predicting user clicks, or enhancing recommendations. In this work, we present a method called HypTrails for comparing a set of hypotheses about human trails on the Web, where hypotheses represent beliefs about transitions between states. Our method utilizes Markov chain models with Bayesian inference. The main idea is to incorporate hypotheses as informative Dirichlet priors and to calculate the evidence of the data under them. For eliciting Dirichlet priors from hypotheses, we present an adaption of the so-called (trial) roulette method, and to compare the relative plausibility of hypotheses, we employ Bayes factors. We demonstrate the general mechanics and applicability of HypTrails by performing experiments with (i) synthetic trails for which we control the mechanisms that have produced them and (ii) empirical trails stemming from different domains including Web site navigation, business reviews, and online music played. Our work expands the repertoire of methods available for studying human trails.

  @article{singer2017bayesian, abstract = {When users interact with the Web today, they leave sequential digital trails on a massive scale. Examples of such human trails include Web navigation, sequences of online restaurant reviews, or online music play lists. Understanding the factors that drive the production of these trails can be useful, for example, for improving underlying network structures, predicting user clicks, or enhancing recommendations. In this work, we present a method called HypTrails for comparing a set of hypotheses about human trails on the Web, where hypotheses represent beliefs about transitions between states. Our method utilizes Markov chain models with Bayesian inference. The main idea is to incorporate hypotheses as informative Dirichlet priors and to calculate the evidence of the data under them. For eliciting Dirichlet priors from hypotheses, we present an adaption of the so-called (trial) roulette method, and to compare the relative plausibility of hypotheses, we employ Bayes factors. We demonstrate the general mechanics and applicability of HypTrails by performing experiments with (i) synthetic trails for which we control the mechanisms that have produced them and (ii) empirical trails stemming from different domains including Web site navigation, business reviews, and online music played. Our work expands the repertoire of methods available for studying human trails.}, address = {New York, NY, USA}, author = {Singer, Philipp and Helic, Denis and Hotho, Andreas and Strohmaier, Markus}, journal = {ACM Trans. Web}, keywords = {version}, month = {jun}, number = 3, pages = {14:1–14:29}, publisher = {ACM}, title = {A Bayesian Method for Comparing Hypotheses About Human Trails}, volume = 11, year = 2017 }
• 
  Sedentary Behavior among National Elite Rowers during Off-Training—A Pilot Study B. Sperlich; M. Becker; A. Hotho; B. Wallmann-Sperlich; M. Sareban; K. Winkert; J. M. Steinacker; G. Treff in Frontiers in Physiology (2017). 8 655.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  The aim of this pilot study was to analyze the off-training physical activity (PA) profile in national elite German U23 rowers during 31 days of their preparation period. The hours spent in each PA category (i.e. sedentary: <1.5 MET; light physical activity: 1.5–3 MET; moderate physical activity: 3–6 MET and vigorous intense physical activity: >6 MET) were calculated for every valid day (i.e. > 480 min of wear time). The off-training PA during 21 weekdays and 10 weekend days of the final 11-wk preparation period was assessed by a wrist-worn multisensory device (Microsoft Band II (MSBII)). A total of 11 rowers provided valid data (i.e. > 480 min/day) for 11.6 week days and 4.8 weekend days during the 31 days observation period. The average sedentary time was 11.63±1.25 hours per day during the week and 12.49±1.10 hours per day on the weekend, with a tendency to be higher on the weekend compared to weekdays (p = 0.06; d = 0.73). The average time in light, moderate and vigorous PA during the weekdays was 1.27±1.15, 0.76±0.37, 0.51±0.44 hours per day and 0.67±0.43, 0.59±0.37, 0.53±0.32 hours per weekend day. Light physical activity was higher during weekdays compared to the weekend (p = 0.04; d = 0.69) Based on our pilot study of eleven national elite rowers we conclude that rowers display a considerable sedentary off-training behavior of more than 11.5 hours/day.

  @article{sperlich2017sedentary, abstract = {The aim of this pilot study was to analyze the off-training physical activity (PA) profile in national elite German U23 rowers during 31 days of their preparation period. The hours spent in each PA category (i.e. sedentary: <1.5 MET; light physical activity: 1.5–3 MET; moderate physical activity: 3–6 MET and vigorous intense physical activity: >6 MET) were calculated for every valid day (i.e. > 480 min of wear time). The off-training PA during 21 weekdays and 10 weekend days of the final 11-wk preparation period was assessed by a wrist-worn multisensory device (Microsoft Band II (MSBII)). A total of 11 rowers provided valid data (i.e. > 480 min/day) for 11.6 week days and 4.8 weekend days during the 31 days observation period. The average sedentary time was 11.63±1.25 hours per day during the week and 12.49±1.10 hours per day on the weekend, with a tendency to be higher on the weekend compared to weekdays (p = 0.06; d = 0.73). The average time in light, moderate and vigorous PA during the weekdays was 1.27±1.15, 0.76±0.37, 0.51±0.44 hours per day and 0.67±0.43, 0.59±0.37, 0.53±0.32 hours per weekend day. Light physical activity was higher during weekdays compared to the weekend (p = 0.04; d = 0.69) Based on our pilot study of eleven national elite rowers we conclude that rowers display a considerable sedentary off-training behavior of more than 11.5 hours/day.}, author = {Sperlich, Billy and Becker, Martin and Hotho, Andreas and Wallmann-Sperlich, Birgit and Sareban, Mahdi and Winkert, Kay and Steinacker, Jürgen M. and Treff, Gunnar}, journal = {Frontiers in Physiology}, keywords = {wearable}, pages = 655, title = {Sedentary Behavior among National Elite Rowers during Off-Training—A Pilot Study}, volume = 8, year = 2017 }
• 
  IP2Vec: Learning Similarities Between IP Addresses M. Ring; D. Landes; A. Dallmann; A. Hotho in 2017 IEEE International Conference on Data Mining Workshops (ICDMW) (2017). 657–666.
  • [ BibTeX ]
  • [ DOI ]
  • [ Download ]
  @article{ring2017ip2vec, author = {Ring, Markus and Landes, Dieter and Dallmann, Alexander and Hotho, Andreas}, journal = {2017 IEEE International Conference on Data Mining Workshops (ICDMW)}, keywords = {similarity}, pages = {657-666}, title = {IP2Vec: Learning Similarities Between IP Addresses}, type = {Publication}, year = 2017 }
• 
  Comparing Hypotheses About Sequential Data: A Bayesian Approach and Its Applications F. Lemmerich; P. Singer; M. Becker; L. Espin-Noboa; D. Dimitrov; D. Helic; A. Hotho; M. Strohmaier Y. Altun, K. Das, T. Mielik{\"a}inen, D. Malerba, J. Stefanowski, J. Read, M. {\v{Z}}itnik, M. Ceci, S. D{\v{z}}eroski (Eds.) (2017). 354–357.
  • [ Abstract ]
  • [ BibTeX ]
  • [ Download ]
  Sequential data can be found in many settings, e.g., as sequences of visited websites or as location sequences of travellers. To improve the understanding of the underlying mechanisms that generate such sequences, the HypTrails approach provides for a novel data analysis method. Based on first-order Markov chain models and Bayesian hypothesis testing, it allows for comparing a set of hypotheses, i.e., beliefs about transitions between states, with respect to their plausibility considering observed data. HypTrails has been successfully employed to study phenomena in the online and the offline world. In this talk, we want to give an introduction to HypTrails and showcase selected real-world applications on urban mobility and reading behavior on Wikipedia.

  @inproceedings{10.1007/978-3-319-71273-4_30, abstract = {Sequential data can be found in many settings, e.g., as sequences of visited websites or as location sequences of travellers. To improve the understanding of the underlying mechanisms that generate such sequences, the HypTrails approach provides for a novel data analysis method. Based on first-order Markov chain models and Bayesian hypothesis testing, it allows for comparing a set of hypotheses, i.e., beliefs about transitions between states, with respect to their plausibility considering observed data. HypTrails has been successfully employed to study phenomena in the online and the offline world. In this talk, we want to give an introduction to HypTrails and showcase selected real-world applications on urban mobility and reading behavior on Wikipedia.}, address = {Cham}, author = {Lemmerich, Florian and Singer, Philipp and Becker, Martin and Espin-Noboa, Lisette and Dimitrov, Dimitar and Helic, Denis and Hotho, Andreas and Strohmaier, Markus}, booktitle = {Machine Learning and Knowledge Discovery in Databases}, editor = {Altun, Yasemin and Das, Kamalika and Mielik{\"a}inen, Taneli and Malerba, Donato and Stefanowski, Jerzy and Read, Jesse and {\v{Z}}itnik, Marinka and Ceci, Michelangelo and D{\v{z}}eroski, Sa{\v{s}}o}, keywords = {sequential}, pages = {354–357}, publisher = {Springer International Publishing}, title = {Comparing Hypotheses About Sequential Data: A Bayesian Approach and Its Applications}, year = 2017 }
• 
  Flow-based benchmark data sets for intrusion detection M. Ring; S. Wunderlich; D. Gr{\"u}dl; D. Landes; A. Hotho (2017). 361–369.
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{ring2017flow, author = {Ring, Markus and Wunderlich, Sarah and Gr{\"u}dl, Dominik and Landes, Dieter and Hotho, Andreas}, booktitle = {Proceedings of the 16th European Conference on Cyber Warfare and Security}, keywords = {citedby:scholar:timestamp:2018-1-31}, pages = {361–369}, title = {Flow-based benchmark data sets for intrusion detection}, year = 2017 }
• 
  Creation of Flow-Based Data Sets for Intrusion Detection M. Ring; S. Wunderlich; D. Grüdl; D. Landes; A. Hotho in Journal of Information Warfare (2017). 16(4) 41–54.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Publicly available labelled data sets are necessary for evaluating anomaly-based Intrusion Detection Systems (IDS). However, existing data sets are often not up-to-date or not yet published because of privacy concerns. This paper identifies requirements for good data sets and proposes an approach for their generation. The key idea is to use a test environment and emulate realistic user behaviour with parameterised scripts on the clients. Comprehensive logging mechanisms provide additional information which may be used for a better understanding of the inner dynamics of an IDS. Finally, the proposed approach is used to generate the flow-based CIDDS-002 data set.

  @article{ring2017creation, abstract = {Publicly available labelled data sets are necessary for evaluating anomaly-based Intrusion Detection Systems (IDS). However, existing data sets are often not up-to-date or not yet published because of privacy concerns. This paper identifies requirements for good data sets and proposes an approach for their generation. The key idea is to use a test environment and emulate realistic user behaviour with parameterised scripts on the clients. Comprehensive logging mechanisms provide additional information which may be used for a better understanding of the inner dynamics of an IDS. Finally, the proposed approach is used to generate the flow-based CIDDS-002 data set.}, author = {Ring, Markus and Wunderlich, Sarah and Grüdl, Dominik and Landes, Dieter and Hotho, Andreas}, journal = {Journal of Information Warfare}, keywords = {myown}, number = 4, pages = {41-54}, title = {Creation of Flow-Based Data Sets for Intrusion Detection}, volume = 16, year = 2017 }
• 
  Improving Session Recommendation with Recurrent Neural Networks by Exploiting Dwell Time A. Dallmann; A. Grimm; C. Pölitz; D. Zoller; A. Hotho (2017).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Recently, Recurrent Neural Networks (RNNs) have been applied to the task of session-based recommendation. These approaches use RNNs to predict the next item in a user session based on the previ- ously visited items. While some approaches consider additional item properties, we argue that item dwell time can be used as an implicit measure of user interest to improve session-based item recommen- dations. We propose an extension to existing RNN approaches that captures user dwell time in addition to the visited items and show that recommendation performance can be improved. Additionally, we investigate the usefulness of a single validation split for model selection in the case of minor improvements and find that in our case the best model is not selected and a fold-like study with different validation sets is necessary to ensure the selection of the best model.

  @misc{dallmann2017improving, abstract = {Recently, Recurrent Neural Networks (RNNs) have been applied to the task of session-based recommendation. These approaches use RNNs to predict the next item in a user session based on the previ- ously visited items. While some approaches consider additional item properties, we argue that item dwell time can be used as an implicit measure of user interest to improve session-based item recommen- dations. We propose an extension to existing RNN approaches that captures user dwell time in addition to the visited items and show that recommendation performance can be improved. Additionally, we investigate the usefulness of a single validation split for model selection in the case of minor improvements and find that in our case the best model is not selected and a fold-like study with different validation sets is necessary to ensure the selection of the best model.}, author = {Dallmann, Alexander and Grimm, Alexander and Pölitz, Christian and Zoller, Daniel and Hotho, Andreas}, keywords = {recommender}, note = {cite arxiv:1706.10231Comment: 6 pages, 3 figures, submission to DLRS workshop}, title = {Improving Session Recommendation with Recurrent Neural Networks by Exploiting Dwell Time}, year = 2017 }
• 
  Towards Sentiment Analysis on German Literature A. Zehe; M. Becker; F. Jannidis; A. Hotho G. Kern-Isberner, J. F{\"u}rnkranz, M. Thimm (Eds.) (2017). 387–394.
  • [ Abstract ]
  • [ BibTeX ]
  • [ Download ]
  Sentiment Analysis is a Natural Language Processing-task that is relevant in a number of contexts, including the analysis of literature. We report on ongoing research towards enabling, for the first time, sentence-level Sentiment Analysis in the domain of German novels. We create a labelled dataset from sentences extracted from German novels and, by adapting existing sentiment classifiers, reach promising F1-scores of 0.67 for binary polarity classification.

  @inproceedings{zehe2017towards, abstract = {Sentiment Analysis is a Natural Language Processing-task that is relevant in a number of contexts, including the analysis of literature. We report on ongoing research towards enabling, for the first time, sentence-level Sentiment Analysis in the domain of German novels. We create a labelled dataset from sentences extracted from German novels and, by adapting existing sentiment classifiers, reach promising F1-scores of 0.67 for binary polarity classification.}, address = {Cham}, author = {Zehe, Albin and Becker, Martin and Jannidis, Fotis and Hotho, Andreas}, booktitle = {KI 2017: Advances in Artificial Intelligence}, editor = {Kern-Isberner, Gabriele and F{\"u}rnkranz, Johannes and Thimm, Matthias}, keywords = {sentiment}, pages = {387–394}, publisher = {Springer International Publishing}, title = {Towards Sentiment Analysis on German Literature}, year = 2017 }

• 
  Prediction of Happy Endings in German Novels A. Zehe; M. Becker; L. Hettinger; A. Hotho; I. Reger; F. Jannidis in CEUR Workshop Proceedings, P. Cellier, T. Charnois, A. Hotho, S. Matwin, M.-F. Moens, Y. Toussaint (Eds.) (2016). (Vol. 1646) 9–16.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{zehe2016prediction, author = {Zehe, Albin and Becker, Martin and Hettinger, Lena and Hotho, Andreas and Reger, Isabella and Jannidis, Fotis}, booktitle = {DMNLP@PKDD/ECML}, crossref = {conf/pkdd/2016dmnlp}, editor = {Cellier, Peggy and Charnois, Thierry and Hotho, Andreas and Matwin, Stan and Moens, Marie-Francine and Toussaint, Yannick}, keywords = {text}, pages = {9-16}, publisher = {CEUR-WS.org}, series = {CEUR Workshop Proceedings}, title = {Prediction of Happy Endings in German Novels}, volume = 1646, year = 2016 }
• 
  Posted, Visited, Exported: Altmetrics in the Social Tagging System BibSonomy D. Zoller; S. Doerfel; R. Jäschke; G. Stumme; A. Hotho in Journal of Informetrics (2016). 10(3) 732–749.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  In social tagging systems, like Mendeley, CiteULike, and BibSonomy, users can post, tag, visit, or export scholarly publications. In this paper, we compare citations with metrics derived from users’ activities (altmetrics) in the popular social bookmarking system BibSonomy. Our analysis, using a corpus of more than 250,000 publications published before 2010, reveals that overall, citations and altmetrics in BibSonomy are mildly correlated. Furthermore, grouping publications by user-generated tags results in topic-homogeneous subsets that exhibit higher correlations with citations than the full corpus. We find that posts, exports, and visits of publications are correlated with citations and even bear predictive power over future impact. Machine learning classifiers predict whether the number of citations that a publication receives in a year exceeds the median number of citations in that year, based on the usage counts of the preceding year. In that setup, a Random Forest predictor outperforms the baseline on average by seven percentage points.

  @article{zoller2016posted, abstract = {In social tagging systems, like Mendeley, CiteULike, and BibSonomy, users can post, tag, visit, or export scholarly publications. In this paper, we compare citations with metrics derived from users’ activities (altmetrics) in the popular social bookmarking system BibSonomy. Our analysis, using a corpus of more than 250,000 publications published before 2010, reveals that overall, citations and altmetrics in BibSonomy are mildly correlated. Furthermore, grouping publications by user-generated tags results in topic-homogeneous subsets that exhibit higher correlations with citations than the full corpus. We find that posts, exports, and visits of publications are correlated with citations and even bear predictive power over future impact. Machine learning classifiers predict whether the number of citations that a publication receives in a year exceeds the median number of citations in that year, based on the usage counts of the preceding year. In that setup, a Random Forest predictor outperforms the baseline on average by seven percentage points.}, author = {Zoller, Daniel and Doerfel, Stephan and Jäschke, Robert and Stumme, Gerd and Hotho, Andreas}, journal = {Journal of Informetrics}, keywords = {scientometrics}, number = 3, pages = {732 - 749}, title = {Posted, Visited, Exported: Altmetrics in the Social Tagging System BibSonomy}, volume = 10, year = 2016 }
• 
  What Users Actually do in a Social Tagging System: A Study of User Behavior in BibSonomy S. Doerfel; D. Zoller; P. Singer; T. Niebler; A. Hotho; M. Strohmaier in ACM Transactions on the Web (2016). 10(2) 14:1–14:32.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Social tagging systems have established themselves as an important part in today’s web and have attracted the interest of our research community in a variety of investigations. Henceforth, several aspects of social tagging systems have been discussed and assumptions have emerged on which our community builds their work. Yet, testing such assumptions has been difficult due to the absence of suitable usage data in the past. In this work, we thoroughly investigate and evaluate four aspects about tagging systems, covering social interaction, retrieval of posted resources, the importance of the three different types of entities, users, resources, and tags, as well as connections between these entities’ popularity in posted and in requested content. For that purpose, we examine live server log data gathered from the real-world, public social tagging system BibSonomy. Our empirical results paint a mixed picture about the four aspects. While for some, typical assumptions hold to a certain extent, other aspects need to be reflected in a very critical light. Our observations have implications for the understanding of social tagging systems, and the way they are used on the web. We make the dataset used in this work available to other researchers.

  @article{doerfel2016users, abstract = {Social tagging systems have established themselves as an important part in today’s web and have attracted the interest of our research community in a variety of investigations. Henceforth, several aspects of social tagging systems have been discussed and assumptions have emerged on which our community builds their work. Yet, testing such assumptions has been difficult due to the absence of suitable usage data in the past. In this work, we thoroughly investigate and evaluate four aspects about tagging systems, covering social interaction, retrieval of posted resources, the importance of the three different types of entities, users, resources, and tags, as well as connections between these entities’ popularity in posted and in requested content. For that purpose, we examine live server log data gathered from the real-world, public social tagging system BibSonomy. Our empirical results paint a mixed picture about the four aspects. While for some, typical assumptions hold to a certain extent, other aspects need to be reflected in a very critical light. Our observations have implications for the understanding of social tagging systems, and the way they are used on the web. We make the dataset used in this work available to other researchers.}, author = {Doerfel, Stephan and Zoller, Daniel and Singer, Philipp and Niebler, Thomas and Hotho, Andreas and Strohmaier, Markus}, journal = {ACM Transactions on the Web}, keywords = {weblog}, month = {may}, number = 2, pages = {14:1–14:32}, publisher = {ACM}, title = {What Users Actually do in a Social Tagging System: A Study of User Behavior in BibSonomy}, volume = 10, year = 2016 }
• 
  Significance Testing for the Classification of Literary Subgenres L. Hettinger; F. Jannidis; I. Reger; A. Hotho (2016).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{hettinger2016significance, author = {Hettinger, Lena and Jannidis, Fotis and Reger, Isabella and Hotho, Andreas}, booktitle = {DH 2016}, keywords = {statistics}, title = {Significance Testing for the Classification of Literary Subgenres}, year = 2016 }
• 
  Classification of Literary Subgenres L. Hettinger; F. Jannidis; I. Reger; A. Hotho (2016).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{hettinger2016classification, author = {Hettinger, Lena and Jannidis, Fotis and Reger, Isabella and Hotho, Andreas}, booktitle = {DHd 2016}, keywords = {novels}, title = {Classification of Literary Subgenres}, year = 2016 }
• 
  Extracting Semantics from Random Walks on Wikipedia: Comparing learning and counting methods A. Dallmann; T. Niebler; F. Lemmerich; A. Hotho R. West, L. Zia, D. Taraborelli, J. Leskovec (Eds.) (2016).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Semantic relatedness between words has been extracted from a variety of sources. In this ongoing work, we explore and compare several options for determining if semantic relatedness can be extracted from navigation structures in Wikipedia. In that direction, we first investigate the potential of representation learning techniques such as DeepWalk in comparison to previously applied methods based on counting co-occurrences. Since both methods are based on (random) paths in the network, we also study different approaches to generate paths from Wikipedia link structure. For this task, we do not only consider the link structure of Wikipedia, but also actual navigation behavior of users. Finally, we analyze if semantics can also be extracted from smaller subsets of the Wikipedia link network. As a result we find that representa- tion learning techniques mostly outperform the investigated co-occurrence counting methods on the Wikipedia network. However, we find that this is not the case for paths sampled from human navigation behavior.

  @inproceedings{dallmann2016extracting, abstract = {Semantic relatedness between words has been extracted from a variety of sources. In this ongoing work, we explore and compare several options for determining if semantic relatedness can be extracted from navigation structures in Wikipedia. In that direction, we first investigate the potential of representation learning techniques such as DeepWalk in comparison to previously applied methods based on counting co-occurrences. Since both methods are based on (random) paths in the network, we also study different approaches to generate paths from Wikipedia link structure. For this task, we do not only consider the link structure of Wikipedia, but also actual navigation behavior of users. Finally, we analyze if semantics can also be extracted from smaller subsets of the Wikipedia link network. As a result we find that representa- tion learning techniques mostly outperform the investigated co-occurrence counting methods on the Wikipedia network. However, we find that this is not the case for paths sampled from human navigation behavior.}, author = {Dallmann, Alexander and Niebler, Thomas and Lemmerich, Florian and Hotho, Andreas}, booktitle = {Wiki Workshop@ICWSM}, editor = {West, Robert and Zia, Leila and Taraborelli, Dario and Leskovec, Jure}, keywords = {word2vec}, title = {Extracting Semantics from Random Walks on Wikipedia: Comparing learning and counting methods}, year = 2016 }
• 
  Proceedings of the Workshop on Interactions between Data Mining and Natural Language Processing, DMNLP 2016, co-located with the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2016, Riva del Garda, Italy, September 23, 2016. P. Cellier; T. Charnois; A. Hotho; S. Matwin; M.-F. Moens; Y. Toussaint in CEUR Workshop Proceedings (2016). (Vol. 1646) CEUR-WS.org.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{conf/pkdd/2016dmnlp, booktitle = {DMNLP@PKDD/ECML}, editor = {Cellier, Peggy and Charnois, Thierry and Hotho, Andreas and Matwin, Stan and Moens, Marie-Francine and Toussaint, Yannick}, keywords = {workshop}, publisher = {CEUR-WS.org}, series = {CEUR Workshop Proceedings}, title = {Proceedings of the Workshop on Interactions between Data Mining and Natural Language Processing, DMNLP 2016, co-located with the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2016, Riva del Garda, Italy, September 23, 2016.}, volume = 1646, year = 2016 }
• 
  Comparison of non-invasive individual monitoring of the training and health of athletes with commercially available wearable technologies P. Düking; A. Hotho; F. K. Fuss; H.-C. Holmberg; B. Sperlich in Frontiers in Physiology (2016). 7(71)
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Athletes adapt their training daily to optimize performance, as well as avoid fatigue, overtraining and other undesirable effects on their health. To optimize training load, each athlete must take his/her own personal objective and subjective characteristics into consideration and an increasing number of wearable technologies (wearables) provide convenient monitoring of various parameters. Accordingly, it is important to help athletes decide which parameters are of primary interest and which wearables can monitor these parameters most effectively. Here, we discuss the wearable technologies available for non-invasive monitoring of various parameters concerning an athlete's training and health. On the basis of these considerations, we suggest directions for future development. Furthermore, we propose that a combination of several wearables is most effective for accessing all relevant parameters, disturbing the athlete as little as possible, and optimizing performance and promoting health.

  @article{10.3389/fphys.2016.00071, abstract = {Athletes adapt their training daily to optimize performance, as well as avoid fatigue, overtraining and other undesirable effects on their health. To optimize training load, each athlete must take his/her own personal objective and subjective characteristics into consideration and an increasing number of wearable technologies (wearables) provide convenient monitoring of various parameters. Accordingly, it is important to help athletes decide which parameters are of primary interest and which wearables can monitor these parameters most effectively. Here, we discuss the wearable technologies available for non-invasive monitoring of various parameters concerning an athlete's training and health. On the basis of these considerations, we suggest directions for future development. Furthermore, we propose that a combination of several wearables is most effective for accessing all relevant parameters, disturbing the athlete as little as possible, and optimizing performance and promoting health.}, author = {Düking, Peter and Hotho, Andreas and Fuss, Franz Konstantin and Holmberg, Hans-Christer and Sperlich, Billy}, journal = {Frontiers in Physiology}, keywords = {wearable}, number = 71, title = {Comparison of non-invasive individual monitoring of the training and health of athletes with commercially available wearable technologies}, volume = 7, year = 2016 }
• 
  Mining Subgroups with Exceptional Transition Behavior F. Lemmerich; M. Becker; P. Singer; D. Helic; A. Hotho; M. Strohmaier B. Krishnapuram, M. Shah, A. J. Smola, C. Aggarwal, D. Shen, R. Rastogi (Eds.) (2016). 965–974.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{lemmerich2016mining, author = {Lemmerich, Florian and Becker, Martin and Singer, Philipp and Helic, Denis and Hotho, Andreas and Strohmaier, Markus}, booktitle = {KDD}, crossref = {conf/kdd/2016}, editor = {Krishnapuram, Balaji and Shah, Mohak and Smola, Alexander J. and Aggarwal, Charu and Shen, Dou and Rastogi, Rajeev}, keywords = {subgroups}, pages = {965-974}, publisher = {ACM}, title = {Mining Subgroups with Exceptional Transition Behavior}, year = 2016 }
• 
  SparkTrails: A MapReduce Implementation of HypTrails for Comparing Hypotheses About Human Trails. M. Becker; H. Mewes; A. Hotho; D. Dimitrov; F. Lemmerich; M. Strohmaier J. Bourdeau, J. Hendler, R. Nkambou, I. Horrocks, B. Y. Zhao (Eds.) (2016). 17–18.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{conf/www/0003MHDLS16, author = {Becker, Martin and Mewes, Hauke and Hotho, Andreas and Dimitrov, Dimitar and Lemmerich, Florian and Strohmaier, Markus}, booktitle = {WWW (Companion Volume)}, crossref = {conf/www/2016c}, editor = {Bourdeau, Jacqueline and Hendler, Jim and Nkambou, Roger and Horrocks, Ian and Zhao, Ben Y.}, keywords = {sparktrails}, pages = {17-18}, publisher = {ACM}, title = {SparkTrails: A MapReduce Implementation of HypTrails for Comparing Hypotheses About Human Trails.}, year = 2016 }
• 
  Analyzing Features for the Detection of Happy Endings in German Novels F. Jannidis; I. Reger; A. Zehe; M. Becker; L. Hettinger; A. Hotho (2016).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  With regard to a computational representation of literary plot, this paper looks at the use of sentiment analysis for happy ending detection in German novels. Its focus lies on the investigation of previously proposed sentiment features in order to gain insight about the relevance of specific features on the one hand and the implications of their performance on the other hand. Therefore, we study various partitionings of novels, considering the highly variable concept of "ending". We also show that our approach, even though still rather simple, can potentially lead to substantial findings relevant to literary studies.

  @misc{jannidis2016analyzing, abstract = {With regard to a computational representation of literary plot, this paper looks at the use of sentiment analysis for happy ending detection in German novels. Its focus lies on the investigation of previously proposed sentiment features in order to gain insight about the relevance of specific features on the one hand and the implications of their performance on the other hand. Therefore, we study various partitionings of novels, considering the highly variable concept of "ending". We also show that our approach, even though still rather simple, can potentially lead to substantial findings relevant to literary studies.}, author = {Jannidis, Fotis and Reger, Isabella and Zehe, Albin and Becker, Martin and Hettinger, Lena and Hotho, Andreas}, keywords = {novel}, note = {cite arxiv:1611.09028}, title = {Analyzing Features for the Detection of Happy Endings in German Novels}, year = 2016 }
• 
  FolkTrails: Interpreting Navigation Behavior in a Social Tagging System T. Niebler; M. Becker; D. Zoller; S. Doerfel; A. Hotho in CIKM ’16 (2016).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Social tagging systems have established themselves as a quick and easy way to organize information by annotating resources with tags. In recent work, user behavior in social tagging systems was studied, that is, how users assign tags, and consume content. However, it is still unclear how users make use of the navigation options they are given. Understanding their behavior and differences in behavior of different user groups is an important step towards assessing the effectiveness of a navigational concept and of improving it to better suit the users’ needs. In this work, we investigate navigation trails in the popular scholarly social tagging system BibSonomy from six years of log data. We discuss dynamic browsing behavior of the general user population and show that different navigational subgroups exhibit different navigational traits. Furthermore, we provide strong evidence that the semantic nature of the underlying folksonomy is an essential factor for explaining navigation.

  @inproceedings{niebler2016folktrails, abstract = {Social tagging systems have established themselves as a quick and easy way to organize information by annotating resources with tags. In recent work, user behavior in social tagging systems was studied, that is, how users assign tags, and consume content. However, it is still unclear how users make use of the navigation options they are given. Understanding their behavior and differences in behavior of different user groups is an important step towards assessing the effectiveness of a navigational concept and of improving it to better suit the users’ needs. In this work, we investigate navigation trails in the popular scholarly social tagging system BibSonomy from six years of log data. We discuss dynamic browsing behavior of the general user population and show that different navigational subgroups exhibit different navigational traits. Furthermore, we provide strong evidence that the semantic nature of the underlying folksonomy is an essential factor for explaining navigation.}, address = {New York, NY, USA}, author = {Niebler, Thomas and Becker, Martin and Zoller, Daniel and Doerfel, Stephan and Hotho, Andreas}, booktitle = {Proceedings of the 25th ACM International on Conference on Information and Knowledge Management}, keywords = {tagging}, publisher = {ACM}, series = {CIKM '16}, title = {FolkTrails: Interpreting Navigation Behavior in a Social Tagging System}, year = 2016 }
• 
  Creation of Specific Flow-Based Training Data Sets for Usage Behaviour Classification F. Otto; M. Ring; D. Landes; A. Hotho (2016). 437.
  • [ BibTeX ]
  @inproceedings{otto2016creation, author = {Otto, Florian and Ring, Markus and Landes, Dieter and Hotho, Andreas}, booktitle = {ECCWS2016-Proceedings fo the 15th European Conference on Cyber Warfare and Security}, keywords = {network}, organization = {Academic Conferences and publishing limited}, pages = 437, title = {Creation of Specific Flow-Based Training Data Sets for Usage Behaviour Classification}, year = 2016 }
• 
  Extracting Semantics from Unconstrained Navigation on Wikipedia T. Niebler; D. Schlör; M. Becker; A. Hotho in KI (2016). 30(2) 163–168.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @article{journals/ki/NieblerS0H16, author = {Niebler, Thomas and Schlör, Daniel and Becker, Martin and Hotho, Andreas}, journal = {KI}, keywords = {wikipedia}, number = 2, pages = {163-168}, title = {Extracting Semantics from Unconstrained Navigation on Wikipedia}, volume = 30, year = 2016 }

• 
  Participatory Patterns in an International Air Quality Monitoring Initiative A. Sîrbu; M. Becker; S. Caminiti; B. De Baets; B. Elen; L. Francis; P. Gravino; A. Hotho; S. Ingarra; V. Loreto; A. Molino; J. Mueller; J. Peters; F. Ricchiuti; F. Saracino; V. D. P. Servedio; G. Stumme; J. Theunis; F. Tria; J. Van den Bossche in PLoS ONE (2015). 10(8) e0136763.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  <p>The issue of sustainability is at the top of the political and societal agenda, being considered of extreme importance and urgency. Human individual action impacts the environment both locally (e.g., local air/water quality, noise disturbance) and globally (e.g., climate change, resource use). Urban environments represent a crucial example, with an increasing realization that the most effective way of producing a change is involving the citizens themselves in monitoring campaigns (a citizen science bottom-up approach). This is possible by developing novel technologies and IT infrastructures enabling large citizen participation. Here, in the wider framework of one of the first such projects, we show results from an international competition where citizens were involved in mobile air pollution monitoring using low cost sensing devices, combined with a web-based game to monitor perceived levels of pollution. Measures of shift in perceptions over the course of the campaign are provided, together with insights into participatory patterns emerging from this study. Interesting effects related to inertia and to direct involvement in measurement activities rather than indirect information exposure are also highlighted, indicating that direct involvement can enhance learning and environmental awareness. In the future, this could result in better adoption of policies towards decreasing pollution.</p>

  @article{10.1371/journal.pone.0136763, abstract = {<p>The issue of sustainability is at the top of the political and societal agenda, being considered of extreme importance and urgency. Human individual action impacts the environment both locally (e.g., local air/water quality, noise disturbance) and globally (e.g., climate change, resource use). Urban environments represent a crucial example, with an increasing realization that the most effective way of producing a change is involving the citizens themselves in monitoring campaigns (a citizen science bottom-up approach). This is possible by developing novel technologies and IT infrastructures enabling large citizen participation. Here, in the wider framework of one of the first such projects, we show results from an international competition where citizens were involved in mobile air pollution monitoring using low cost sensing devices, combined with a web-based game to monitor perceived levels of pollution. Measures of shift in perceptions over the course of the campaign are provided, together with insights into participatory patterns emerging from this study. Interesting effects related to inertia and to direct involvement in measurement activities rather than indirect information exposure are also highlighted, indicating that direct involvement can enhance learning and environmental awareness. In the future, this could result in better adoption of policies towards decreasing pollution.</p>}, author = {Sîrbu, Alina and Becker, Martin and Caminiti, Saverio and De Baets, Bernard and Elen, Bart and Francis, Louise and Gravino, Pietro and Hotho, Andreas and Ingarra, Stefano and Loreto, Vittorio and Molino, Andrea and Mueller, Juergen and Peters, Jan and Ricchiuti, Ferdinando and Saracino, Fabio and Servedio, Vito D. P. and Stumme, Gerd and Theunis, Jan and Tria, Francesca and Van den Bossche, Joris}, journal = {PLoS ONE}, keywords = {selected}, month = {08}, number = 8, pages = {e0136763}, publisher = {Public Library of Science}, title = {Participatory Patterns in an International Air Quality Monitoring Initiative}, volume = 10, year = 2015 }
• 
  Hyptrails: A bayesian approach for comparing hypotheses about human trails P. Singer; D. Helic; A. Hotho; M. Strohmaier (2015).
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{singer2015hyptrails, address = {Firenze, Italy}, author = {Singer, P. and Helic, D. and Hotho, A. and Strohmaier, M.}, booktitle = {24th International World Wide Web Conference (WWW2015)}, keywords = {web}, month = {May 18 - May 22}, organization = {ACM}, publisher = {ACM}, title = {Hyptrails: A bayesian approach for comparing hypotheses about human trails}, year = 2015 }
• 
  Media Bias in German Online Newspapers A. Dallmann; F. Lemmerich; D. Zoller; A. Hotho (2015).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{dallmann2015media, address = {Cyprus, Turkey, September 1-4}, author = {Dallmann, Alexander and Lemmerich, Florian and Zoller, Daniel and Hotho, Andreas}, booktitle = {26th ACM Conference on Hypertext and Social Media}, keywords = {newspaper}, publisher = {ACM}, title = {Media Bias in German Online Newspapers}, year = 2015 }
• 
  ConDist: A Context-Driven Categorical Distance Measure M. Ring; F. Otto; M. Becker; T. Niebler; D. Landes; A. Hotho ECMLPKDD2015 (Ed.) (2015).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{ring2015condist, author = {Ring, Markus and Otto, Florian and Becker, Martin and Niebler, Thomas and Landes, Dieter and Hotho, Andreas}, editor = {ECMLPKDD2015}, keywords = {unsupervised}, title = {ConDist: A Context-Driven Categorical Distance Measure}, year = 2015 }
• 
  Modeling and Extracting Load Intensity Profiles J. v. Kistowski; H. Nikolas; D. Zoller; S. Kounev; A. Hotho (2015).
  • [ Abstract ]
  • [ BibTeX ]
  Today’s system developers and operators face the challenge of creating software systems that make efficient use of dynamically allocated resources under highly variable and dynamic load profiles, while at the same time delivering reliable performance. Benchmarking of systems under these constraints is difficult, as state-of-the-art benchmarking frameworks provide only limited support for emulating such dynamic and highly vari- able load profiles for the creation of realistic workload scenarios. Industrial benchmarks typically confine themselves to workloads with constant or stepwise increasing loads. Alternatively, they support replaying of recorded load traces. Statistical load inten- sity descriptions also do not sufficiently capture concrete pattern load profile variations over time. To address these issues, we present the Descartes Load Intensity Model (DLIM). DLIM provides a modeling formalism for describing load intensity variations over time. A DLIM instance can be used as a compact representation of a recorded load intensity trace, providing a powerful tool for benchmarking and performance analysis. As manually obtaining DLIM instances can be time consuming, we present three different automated extraction methods, which also help to enable autonomous system analysis for self-adaptive systems. Model expressiveness is validated using the presented extraction methods. Extracted DLIM instances exhibit a median modeling error of 12.4% on average over nine different real-world traces covering between two weeks and seven months. Additionally, extraction methods perform orders of magnitude faster than existing time series decomposition approaches.

  @inproceedings{vkistowski2015modeling, abstract = {Today’s system developers and operators face the challenge of creating software systems that make efficient use of dynamically allocated resources under highly variable and dynamic load profiles, while at the same time delivering reliable performance. Benchmarking of systems under these constraints is difficult, as state-of-the-art benchmarking frameworks provide only limited support for emulating such dynamic and highly vari- able load profiles for the creation of realistic workload scenarios. Industrial benchmarks typically confine themselves to workloads with constant or stepwise increasing loads. Alternatively, they support replaying of recorded load traces. Statistical load inten- sity descriptions also do not sufficiently capture concrete pattern load profile variations over time. To address these issues, we present the Descartes Load Intensity Model (DLIM). DLIM provides a modeling formalism for describing load intensity variations over time. A DLIM instance can be used as a compact representation of a recorded load intensity trace, providing a powerful tool for benchmarking and performance analysis. As manually obtaining DLIM instances can be time consuming, we present three different automated extraction methods, which also help to enable autonomous system analysis for self-adaptive systems. Model expressiveness is validated using the presented extraction methods. Extracted DLIM instances exhibit a median modeling error of 12.4% on average over nine different real-world traces covering between two weeks and seven months. Additionally, extraction methods perform orders of magnitude faster than existing time series decomposition approaches.}, author = {v. Kistowski, Jóakim and Nikolas, Herbst. and Zoller, Daniel and Kounev, Samuel and Hotho, Andreas}, booktitle = {Proceedings of the 10th International Symposium on Software Engineering for Adaptive and Self-Managing Systems (SEAMS)}, keywords = {profile}, title = {Modeling and Extracting Load Intensity Profiles}, year = 2015 }
• 
  On Publication Usage in a Social Bookmarking System D. Zoller; S. Doerfel; R. Jäschke; G. Stumme; A. Hotho (2015).
  • [ Abstract ]
  • [ BibTeX ]
  Scholarly success is traditionally measured in terms of citations to publications. With the advent of publication man- agement and digital libraries on the web, scholarly usage data has become a target of investigation and new impact metrics computed on such usage data have been proposed – so called altmetrics. In scholarly social bookmarking sys- tems, scientists collect and manage publication meta data and thus reveal their interest in these publications. In this work, we investigate connections between usage metrics and citations, and find posts, exports, and page views of publications to be correlated to citations.

  @inproceedings{zoller2015publication, abstract = {Scholarly success is traditionally measured in terms of citations to publications. With the advent of publication man- agement and digital libraries on the web, scholarly usage data has become a target of investigation and new impact metrics computed on such usage data have been proposed – so called altmetrics. In scholarly social bookmarking sys- tems, scientists collect and manage publication meta data and thus reveal their interest in these publications. In this work, we investigate connections between usage metrics and citations, and find posts, exports, and page views of publications to be correlated to citations.}, author = {Zoller, Daniel and Doerfel, Stephan and Jäschke, Robert and Stumme, Gerd and Hotho, Andreas}, booktitle = {Proceedings of the 2015 ACM Conference on Web Science}, keywords = {usage}, title = {On Publication Usage in a Social Bookmarking System}, year = 2015 }
• 
  Participatory Patterns in an International Air Quality Monitoring Initiative A. Sîrbu; M. Becker; S. Caminiti; B. De Baets; B. Elen; L. Francis; P. Gravino; A. Hotho; S. Ingarra; V. Loreto; A. Molino; J. Mueller; J. Peters; F. Ricchiuti; F. Saracino; V. D. P. Servedio; G. Stumme; J. Theunis; F. Tria; J. Van den Bossche (2015).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  The issue of sustainability is at the top of the political and societal agenda, being considered of extreme importance and urgency. Human individual action impacts the environment both locally (e.g., local air/water quality, noise disturbance) and globally (e.g., climate change, resource use). Urban environments represent a crucial example, with an increasing realization that the most effective way of producing a change is involving the citizens themselves in monitoring campaigns (a citizen science bottom-up approach). This is possible by developing novel technologies and IT infrastructures enabling large citizen participation. Here, in the wider framework of one of the first such projects, we show results from an international competition where citizens were involved in mobile air pollution monitoring using low cost sensing devices, combined with a web-based game to monitor perceived levels of pollution. Measures of shift in perceptions over the course of the campaign are provided, together with insights into participatory patterns emerging from this study. Interesting effects related to inertia and to direct involvement in measurement activities rather than indirect information exposure are also highlighted, indicating that direct involvement can enhance learning and environmental awareness. In the future, this could result in better adoption of policies towards decreasing pollution.

  @misc{sirbu2015participatory, abstract = {The issue of sustainability is at the top of the political and societal agenda, being considered of extreme importance and urgency. Human individual action impacts the environment both locally (e.g., local air/water quality, noise disturbance) and globally (e.g., climate change, resource use). Urban environments represent a crucial example, with an increasing realization that the most effective way of producing a change is involving the citizens themselves in monitoring campaigns (a citizen science bottom-up approach). This is possible by developing novel technologies and IT infrastructures enabling large citizen participation. Here, in the wider framework of one of the first such projects, we show results from an international competition where citizens were involved in mobile air pollution monitoring using low cost sensing devices, combined with a web-based game to monitor perceived levels of pollution. Measures of shift in perceptions over the course of the campaign are provided, together with insights into participatory patterns emerging from this study. Interesting effects related to inertia and to direct involvement in measurement activities rather than indirect information exposure are also highlighted, indicating that direct involvement can enhance learning and environmental awareness. In the future, this could result in better adoption of policies towards decreasing pollution.}, author = {Sîrbu, Alina and Becker, Martin and Caminiti, Saverio and De Baets, Bernard and Elen, Bart and Francis, Louise and Gravino, Pietro and Hotho, Andreas and Ingarra, Stefano and Loreto, Vittorio and Molino, Andrea and Mueller, Juergen and Peters, Jan and Ricchiuti, Ferdinando and Saracino, Fabio and Servedio, Vito D. P. and Stumme, Gerd and Theunis, Jan and Tria, Francesca and Van den Bossche, Joris}, keywords = {myown}, note = {cite arxiv:1503.07730Comment: 17 pages, 6 figures, 1 supplementary file}, title = {Participatory Patterns in an International Air Quality Monitoring Initiative}, year = 2015 }
• 
  Automatic Threshold Calculation for the Categorical Distance Measure ConDist. M. Ring; D. Landes; A. Hotho in CEUR Workshop Proceedings, R. Bergmann, S. Görg, G. Müller (Eds.) (2015). (Vol. 1458) 52–63.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{ring2015automatic, author = {Ring, Markus and Landes, Dieter and Hotho, Andreas}, booktitle = {LWA}, crossref = {conf/lwa/2015}, editor = {Bergmann, Ralph and Görg, Sebastian and Müller, Gilbert}, keywords = {myown}, pages = {52-63}, publisher = {CEUR-WS.org}, series = {CEUR Workshop Proceedings}, title = {Automatic Threshold Calculation for the Categorical Distance Measure ConDist.}, volume = 1458, year = 2015 }
• 
  Photowalking the city: comparing hypotheses about urban photo trails on Flickr M. Becker; P. Singer; F. Lemmerich; A. Hotho; D. Helic; M. Strohmaier (2015).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{becker2015photowalking, address = {Beijing, China}, author = {Becker, Martin and Singer, Philipp and Lemmerich, Florian and Hotho, Andreas and Helic, Denis and Strohmaier, Markus}, booktitle = {Proceedings of the 7th International Conference on SocialInformatics (to appear)}, keywords = {urban}, month = {dec}, title = {Photowalking the city: comparing hypotheses about urban photo trails on Flickr}, type = {Publication}, year = 2015 }
• 
  VizTrails: An Information Visualization Tool for Exploring Geographic Movement Trajectories M. Becker; P. Singer; F. Lemmerich; A. Hotho; D. Helic; M. Strohmaier in HT ’15 (2015). 319–320.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Understanding the way people move through urban areas represents an important problem that has implications for a range of societal challenges such as city planning, public transportation, or crime analysis. In this paper, we present an interactive visualization tool called VizTrails for exploring and understanding such human movement. It features visualizations that show aggregated statistics of trails for geographic areas that correspond to grid cells on a map, e.g., on the number of users passing through or on cells commonly visited next. Amongst other features, system allows to overlay the map with the results of SPARQL queries in order to relate the observed trajectory statistics with its geo-spatial context, e.g., considering a city's points of interest. The systems functionality is demonstrated using trajectory examples extracted from the social photo sharing platform Flickr. Overall, VizTrails facilitates deeper insights into geo-spatial trajectory data by enabling interactive exploration of aggregated statistics and providing geo-spatial context.

  @inproceedings{Becker:2015:VIV:2700171.2791021, abstract = {Understanding the way people move through urban areas represents an important problem that has implications for a range of societal challenges such as city planning, public transportation, or crime analysis. In this paper, we present an interactive visualization tool called VizTrails for exploring and understanding such human movement. It features visualizations that show aggregated statistics of trails for geographic areas that correspond to grid cells on a map, e.g., on the number of users passing through or on cells commonly visited next. Amongst other features, system allows to overlay the map with the results of SPARQL queries in order to relate the observed trajectory statistics with its geo-spatial context, e.g., considering a city's points of interest. The systems functionality is demonstrated using trajectory examples extracted from the social photo sharing platform Flickr. Overall, VizTrails facilitates deeper insights into geo-spatial trajectory data by enabling interactive exploration of aggregated statistics and providing geo-spatial context.}, address = {New York, NY, USA}, author = {Becker, Martin and Singer, Philipp and Lemmerich, Florian and Hotho, Andreas and Helic, Denis and Strohmaier, Markus}, booktitle = {Proceedings of the 26th ACM Conference on Hypertext \&\#38; Social Media}, keywords = {viztails}, pages = {319–320}, publisher = {ACM}, series = {HT '15}, title = {VizTrails: An Information Visualization Tool for Exploring Geographic Movement Trajectories}, year = 2015 }
• 
  Genre classification on German novels L. Hettinger; M. Becker; I. Reger; F. Jannidis; A. Hotho (2015).
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{schwemmlein2015genre, author = {Hettinger, Lena and Becker, Martin and Reger, Isabella and Jannidis, Fotis and Hotho, Andreas}, booktitle = {Proceedings of the 12th International Workshop on Text-based Information Retrieval}, keywords = {text}, title = {Genre classification on German novels}, year = 2015 }
• 
  MicroTrails: Comparing Hypotheses About Task Selection on a Crowdsourcing Platform M. Becker; K. Borchert; M. Hirth; H. Mewes; A. Hotho; P. Tran-Gia in i-KNOW ’15 (2015). 10:1–10:8.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  To optimize the workflow on commercial crowdsourcing platforms like Amazon Mechanical Turk or Microworkers, it is important to understand how users choose their tasks. Current work usually explores the underlying processes by employing user studies based on surveys with a limited set of participants. In contrast, we formulate hypotheses based on the different findings in these studies and, instead of verifying them based on user feedback, we compare them directly on data from a commercial crowdsourcing platform. For evaluation, we use a Bayesian approach called HypTrails which allows us to give a relative ranking of the corresponding hypotheses. The hypotheses considered, are for example based on task categories, monetary incentives or semantic similarity of task descriptions. We find that, in our scenario, hypotheses based on employers as well the the task descriptions work best. Overall, we objectively compare different factors influencing users when choosing their tasks. Our approach enables crowdsourcing companies to better understand their users in order to optimize their platforms, e.g., by incorparting the gained knowledge about these factors into task recommentation systems.

  @inproceedings{becker2015microtrails, abstract = {To optimize the workflow on commercial crowdsourcing platforms like Amazon Mechanical Turk or Microworkers, it is important to understand how users choose their tasks. Current work usually explores the underlying processes by employing user studies based on surveys with a limited set of participants. In contrast, we formulate hypotheses based on the different findings in these studies and, instead of verifying them based on user feedback, we compare them directly on data from a commercial crowdsourcing platform. For evaluation, we use a Bayesian approach called HypTrails which allows us to give a relative ranking of the corresponding hypotheses. The hypotheses considered, are for example based on task categories, monetary incentives or semantic similarity of task descriptions. We find that, in our scenario, hypotheses based on employers as well the the task descriptions work best. Overall, we objectively compare different factors influencing users when choosing their tasks. Our approach enables crowdsourcing companies to better understand their users in order to optimize their platforms, e.g., by incorparting the gained knowledge about these factors into task recommentation systems.}, address = {New York, NY, USA}, author = {Becker, Martin and Borchert, Kathrin and Hirth, Matthias and Mewes, Hauke and Hotho, Andreas and Tran-Gia, Phuoc}, booktitle = {Proceedings of the 15th International Conference on Knowledge Technologies and Data-driven Business}, keywords = {working}, pages = {10:1–10:8}, publisher = {ACM}, series = {i-KNOW '15}, title = {MicroTrails: Comparing Hypotheses About Task Selection on a Crowdsourcing Platform}, year = 2015 }

• 
  Proceedings of the 1st International Workshop on Interactions between Data Mining and Natural Language Processing co-located with The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, DMNLP@PKDD/ECML 2014, Nancy, France, September 15, 2014 P. Cellier; T. Charnois; A. Hotho; S. Matwin; M.- }Francine Moens; Y. Toussaint in {CEUR} Workshop Proceedings (2014). (Vol. 1202) CEUR-WS.org.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @proceedings{cellier2014proceedings, editor = {Cellier, Peggy and Charnois, Thierry and Hotho, Andreas and Matwin, Stan and Moens, Marie{-}Francine and Toussaint, Yannick}, keywords = {workshop}, publisher = {CEUR-WS.org}, series = {{CEUR} Workshop Proceedings}, title = {Proceedings of the 1st International Workshop on Interactions between Data Mining and Natural Language Processing co-located with The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, DMNLP@PKDD/ECML 2014, Nancy, France, September 15, 2014}, volume = 1202, year = 2014 }
• 
  Of course we share! Testing Assumptions about Social Tagging Systems S. Doerfel; D. Zoller; P. Singer; T. Niebler; A. Hotho; M. Strohmaier (2014).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Social tagging systems have established themselves as an important part in today's web and have attracted the interest from our research community in a variety of investigations. The overall vision of our community is that simply through interactions with the system, i.e., through tagging and sharing of resources, users would contribute to building useful semantic structures as well as resource indexes using uncontrolled vocabulary not only due to the easy-to-use mechanics. Henceforth, a variety of assumptions about social tagging systems have emerged, yet testing them has been difficult due to the absence of suitable data. In this work we thoroughly investigate three available assumptions - e.g., is a tagging system really social? - by examining live log data gathered from the real-world public social tagging system BibSonomy. Our empirical results indicate that while some of these assumptions hold to a certain extent, other assumptions need to be reflected and viewed in a very critical light. Our observations have implications for the design of future search and other algorithms to better reflect the actual user behavior.

  @misc{doerfel2014course, abstract = {Social tagging systems have established themselves as an important part in today's web and have attracted the interest from our research community in a variety of investigations. The overall vision of our community is that simply through interactions with the system, i.e., through tagging and sharing of resources, users would contribute to building useful semantic structures as well as resource indexes using uncontrolled vocabulary not only due to the easy-to-use mechanics. Henceforth, a variety of assumptions about social tagging systems have emerged, yet testing them has been difficult due to the absence of suitable data. In this work we thoroughly investigate three available assumptions - e.g., is a tagging system really social? - by examining live log data gathered from the real-world public social tagging system BibSonomy. Our empirical results indicate that while some of these assumptions hold to a certain extent, other assumptions need to be reflected and viewed in a very critical light. Our observations have implications for the design of future search and other algorithms to better reflect the actual user behavior.}, author = {Doerfel, Stephan and Zoller, Daniel and Singer, Philipp and Niebler, Thomas and Hotho, Andreas and Strohmaier, Markus}, keywords = {tagging}, note = {cite arxiv:1401.0629}, title = {Of course we share! Testing Assumptions about Social Tagging Systems}, year = 2014 }
• 
  {Ubicon and its Applications for Ubiquitous Social Computing} M. Atzmueller; M. Becker; M. Kibanov; C. Scholz; S. Doerfel; A. Hotho; B.-E. Macek; F. Mitzlaff; J. Mueller; G. Stumme in New Review of Hypermedia and Multimedia (2014). 20(1) 53–77.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  The combination of ubiquitous and social computing is an emerging research area which integrates different but complementary methods, techniques and tools. In this paper, we focus on the Ubicon platform, its applications, and a large spectrum of analysis results. Ubicon provides an extensible framework for building and hosting applications targeting both ubiquitous and social environments. We summarize the architecture and exemplify its implementation using four real-world applications built on top of Ubicon. In addition, we discuss several scientific experiments in the context of these applications in order to give a better picture of the potential of the framework, and discuss analysis results using several real-world data sets collected utilizing Ubicon.

  @article{atzmueller2014ubicon, abstract = {The combination of ubiquitous and social computing is an emerging research area which integrates different but complementary methods, techniques and tools. In this paper, we focus on the Ubicon platform, its applications, and a large spectrum of analysis results. Ubicon provides an extensible framework for building and hosting applications targeting both ubiquitous and social environments. We summarize the architecture and exemplify its implementation using four real-world applications built on top of Ubicon. In addition, we discuss several scientific experiments in the context of these applications in order to give a better picture of the potential of the framework, and discuss analysis results using several real-world data sets collected utilizing Ubicon.}, author = {Atzmueller, Martin and Becker, Martin and Kibanov, Mark and Scholz, Christoph and Doerfel, Stephan and Hotho, Andreas and Macek, Bjoern-Elmar and Mitzlaff, Folke and Mueller, Juergen and Stumme, Gerd}, journal = {New Review of Hypermedia and Multimedia}, keywords = {ubiquitous}, number = 1, pages = {53–77}, title = {{Ubicon and its Applications for Ubiquitous Social Computing}}, volume = 20, year = 2014 }
• 
  Evaluating Assumptions about Social Tagging - {A} Study of User Behavior in BibSonomy S. Doerfel; D. Zoller; P. Singer; T. Niebler; A. Hotho; M. Strohmaier T. Seidl, M. Hassani, C. Beecks (Eds.) (2014). 18–19.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{doerfel2014evaluating, author = {Doerfel, Stephan and Zoller, Daniel and Singer, Philipp and Niebler, Thomas and Hotho, Andreas and Strohmaier, Markus}, booktitle = {Proceedings of the 16th {LWA} Workshops: KDML, {IR} and FGWM, Aachen, Germany, September 8-10, 2014.}, editor = {Seidl, Thomas and Hassani, Marwan and Beecks, Christian}, keywords = {tagging}, pages = {18–19}, publisher = {CEUR-WS.org}, title = {Evaluating Assumptions about Social Tagging - {A} Study of User Behavior in BibSonomy}, year = 2014 }
• 
  How Social is Social Tagging? S. Doerfel; D. Zoller; P. Singer; T. Niebler; A. Hotho; M. Strohmaier in WWW 2014 (2014).
  • [ BibTeX ]
  • [ Download ]
  @inproceedings{doerfel2014social, address = {New York, NY, USA}, author = {Doerfel, Stephan and Zoller, Daniel and Singer, Philipp and Niebler, Thomas and Hotho, Andreas and Strohmaier, Markus}, booktitle = {Proceedings of the 23rd International World Wide Web Conference}, keywords = {tagging}, publisher = {ACM}, series = {WWW 2014}, title = {How Social is Social Tagging?}, year = 2014 }
• 
  The sixth {ACM} RecSys workshop on recommender systems and the social web D. Jannach; J. Freyne; W. Geyer; I. Guy; A. Hotho; B. Mobasher (2014). 395.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @inproceedings{jannach2014sixth, author = {Jannach, Dietmar and Freyne, Jill and Geyer, Werner and Guy, Ido and Hotho, Andreas and Mobasher, Bamshad}, booktitle = {Eighth {ACM} Conference on Recommender Systems, RecSys '14, Foster City, Silicon Valley, CA, {USA} - October 06 - 10, 2014}, keywords = {workshop}, pages = 395, title = {The sixth {ACM} RecSys workshop on recommender systems and the social web}, year = 2014 }
• 
  HypTrails: A Bayesian Approach for Comparing Hypotheses about Human Trails on the Web P. Singer; D. Helic; A. Hotho; M. Strohmaier (2014).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  When users interact with the Web today, they leave sequential digital trails on a massive scale. Examples of such human trails include Web navigation, sequences of online restaurant reviews, or online music play lists. Understanding the factors that drive the production of these trails can be useful for e.g., improving underlying network structures, predicting user clicks or enhancing recommendations. In this work, we present a general approach called HypTrails for comparing a set of hypotheses about human trails on the Web, where hypotheses represent beliefs about transitions between states. Our approach utilizes Markov chain models with Bayesian inference. The main idea is to incorporate hypotheses as informative Dirichlet priors and to leverage the sensitivity of Bayes factors on the prior for comparing hypotheses with each other. For eliciting Dirichlet priors from hypotheses, we present an adaption of the so-called (trial) roulette method. We demonstrate the general mechanics and applicability of HypTrails by performing experiments with (i) synthetic trails for which we control the mechanisms that have produced them and (ii) empirical trails stemming from different domains including website navigation, business reviews and online music played. Our work expands the repertoire of methods available for studying human trails on the Web.

  @misc{singer2014hyptrails, abstract = {When users interact with the Web today, they leave sequential digital trails on a massive scale. Examples of such human trails include Web navigation, sequences of online restaurant reviews, or online music play lists. Understanding the factors that drive the production of these trails can be useful for e.g., improving underlying network structures, predicting user clicks or enhancing recommendations. In this work, we present a general approach called HypTrails for comparing a set of hypotheses about human trails on the Web, where hypotheses represent beliefs about transitions between states. Our approach utilizes Markov chain models with Bayesian inference. The main idea is to incorporate hypotheses as informative Dirichlet priors and to leverage the sensitivity of Bayes factors on the prior for comparing hypotheses with each other. For eliciting Dirichlet priors from hypotheses, we present an adaption of the so-called (trial) roulette method. We demonstrate the general mechanics and applicability of HypTrails by performing experiments with (i) synthetic trails for which we control the mechanisms that have produced them and (ii) empirical trails stemming from different domains including website navigation, business reviews and online music played. Our work expands the repertoire of methods available for studying human trails on the Web.}, author = {Singer, Philipp and Helic, Denis and Hotho, Andreas and Strohmaier, Markus}, keywords = {sys:relevantfor:l3s}, note = {cite arxiv:1411.2844}, title = {HypTrails: A Bayesian Approach for Comparing Hypotheses about Human Trails on the Web}, year = 2014 }
• 
  Proceedings of the 6th Workshop on Recommender Systems and the Social Web (RSWeb 2014) co-located with the 8th {ACM} Conference on Recommender Systems (RecSys 2014), Foster City, CA, USA, October 6, 2014 D. Jannach; J. Freyne; W. Geyer; I. Guy; A. Hotho; B. Mobasher in {CEUR} Workshop Proceedings (2014). (Vol. 1271) CEUR-WS.org.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{jannach2014proceedings, editor = {Jannach, Dietmar and Freyne, Jill and Geyer, Werner and Guy, Ido and Hotho, Andreas and Mobasher, Bamshad}, keywords = {workshop}, publisher = {CEUR-WS.org}, series = {{CEUR} Workshop Proceedings}, title = {Proceedings of the 6th Workshop on Recommender Systems and the Social Web (RSWeb 2014) co-located with the 8th {ACM} Conference on Recommender Systems (RecSys 2014), Foster City, CA, USA, October 6, 2014}, volume = 1271, year = 2014 }
• 
  The social distributional hypothesis: a pragmatic proxy for homophily in online social networks F. Mitzlaff; M. Atzmueller; A. Hotho; G. Stumme in Social Network Analysis and Mining (2014). 4(1)
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  Applications of the Social Web are ubiquitous and have become an integral part of everyday life: Users make friends, for example, with the help of online social networks, share thoughts via Twitter, or collaboratively write articles in Wikipedia. All such interactions leave digital traces; thus, users participate in the creation of heterogeneous, distributed, collaborative data collections. In linguistics, the

  @article{noKey, abstract = {Applications of the Social Web are ubiquitous and have become an integral part of everyday life: Users make friends, for example, with the help of online social networks, share thoughts via Twitter, or collaboratively write articles in Wikipedia. All such interactions leave digital traces; thus, users participate in the creation of heterogeneous, distributed, collaborative data collections. In linguistics, the}, author = {Mitzlaff, Folke and Atzmueller, Martin and Hotho, Andreas and Stumme, Gerd}, journal = {Social Network Analysis and Mining}, keywords = {sys:relevantfor:l3s}, number = 1, publisher = {Springer Vienna}, title = {The social distributional hypothesis: a pragmatic proxy for homophily in online social networks}, volume = 4, year = 2014 }
• 
  Folksonomies P. Singer; T. Niebler; A. Hotho; M. Strohmaier in Encyclopedia of Social Network Analysis and Mining (2014). 542–547.
  • [ BibTeX ]
  @incollection{singer2014folksonomies, author = {Singer, Philipp and Niebler, Thomas and Hotho, Andreas and Strohmaier, Markus}, booktitle = {Encyclopedia of Social Network Analysis and Mining}, keywords = {myown}, pages = {542–547}, publisher = {Springer}, title = {Folksonomies}, year = 2014 }
• 
  Subjective vs. Objective Data: Bridging the Gap M. Becker; A. Hotho; J. Mueller; M. Kibanov; M. Atzmueller; G. Stumme (2014).
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Sensor data is objective. But when measuring our environment, measured values are contrasted with our perception, which is always subjective. This makes interpreting sensor measurements difficult for a single person in her personal environment. In this context, the EveryAware projects directly connects the concepts of objective sensor data with subjective impressions and perceptions by providing a collective sensing platform with several client applications allowing to explicitly associate those two data types. The goal is to provide the user with personalized feedback, a characterization of the global as well as her personal environment, and enable her to position her perceptions in this global context. In this poster we summarize the collected data of two EveryAware applications, namely WideNoise for noise measurements and AirProbe for participatory air quality sensing. Basic insights are presented including user activity, learning processes and sensor data to perception correlations. These results provide an outlook on how this data can further be used to understand the connection between sensor data and perceptions.

  @misc{becker2014subjective, abstract = {Sensor data is objective. But when measuring our environment, measured values are contrasted with our perception, which is always subjective. This makes interpreting sensor measurements difficult for a single person in her personal environment. In this context, the EveryAware projects directly connects the concepts of objective sensor data with subjective impressions and perceptions by providing a collective sensing platform with several client applications allowing to explicitly associate those two data types. The goal is to provide the user with personalized feedback, a characterization of the global as well as her personal environment, and enable her to position her perceptions in this global context. In this poster we summarize the collected data of two EveryAware applications, namely WideNoise for noise measurements and AirProbe for participatory air quality sensing. Basic insights are presented including user activity, learning processes and sensor data to perception correlations. These results provide an outlook on how this data can further be used to understand the connection between sensor data and perceptions.}, author = {Becker, Martin and Hotho, Andreas and Mueller, Juergen and Kibanov, Mark and Atzmueller, Martin and Stumme, Gerd}, howpublished = {CSSWS 2014, Poster}, keywords = {myown}, title = {Subjective vs. Objective Data: Bridging the Gap}, year = 2014 }

• 
  Informationelle Selbstbestimmung Im Web 2.0 Chancen Und Risiken Sozialer Verschlagwortungssysteme S. Doerfel; A. Hotho; A. Kartal-Aydemir; A. Roßnagel; G. Stumme (2013). Vieweg + Teubner Verlag.
  • [ BibTeX ]
  • [ URL ]
  @book{doerfel2013informationelle, author = {Doerfel, Stephan and Hotho, Andreas and Kartal-Aydemir, Aliye and Roßnagel, Alexander and Stumme, Gerd}, keywords = {sys:relevantfor:l3s}, publisher = {Vieweg + Teubner Verlag}, title = {Informationelle Selbstbestimmung Im Web 2.0 Chancen Und Risiken Sozialer Verschlagwortungssysteme}, year = 2013 }
• 
  How Tagging Pragmatics Influence Tag Sense Discovery in Social Annotation Systems T. Niebler; P. Singer; D. Benz; C. Körner; M. Strohmaier; A. Hotho in Advances in Information Retrieval, P. Serdyukov, P. Braslavski, S. Kuznetsov, J. Kamps, S. Rüger, E. Agichtein, I. Segalovich, E. Yilmaz (Eds.) (2013). (Vol. 7814) 86–97.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  The presence of emergent semantics in social annotation systems has been reported in numerous studies. Two important problems in this context are the induction of semantic relations among tags and the discovery of different senses of a given tag. While a number of approaches for discovering tag senses exist, little is known about which

  @incollection{niebler2013tagging, abstract = {The presence of emergent semantics in social annotation systems has been reported in numerous studies. Two important problems in this context are the induction of semantic relations among tags and the discovery of different senses of a given tag. While a number of approaches for discovering tag senses exist, little is known about which}, author = {Niebler, Thomas and Singer, Philipp and Benz, Dominik and Körner, Christian and Strohmaier, Markus and Hotho, Andreas}, booktitle = {Advances in Information Retrieval}, editor = {Serdyukov, Pavel and Braslavski, Pavel and Kuznetsov, SergeiO. and Kamps, Jaap and Rüger, Stefan and Agichtein, Eugene and Segalovich, Ilya and Yilmaz, Emine}, keywords = {tagging}, pages = {86-97}, publisher = {Springer Berlin Heidelberg}, series = {Lecture Notes in Computer Science}, title = {How Tagging Pragmatics Influence Tag Sense Discovery in Social Annotation Systems}, volume = 7814, year = 2013 }
• 
  Deeper Into the Folksonomy Graph: FolkRank Adaptations and Extensions for Improved Tag Recommendations N. Landia; S. Doerfel; R. Jäschke; S. S. Anand; A. Hotho; N. Griffiths in cs.IR (2013). 1310.1498
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  The information contained in social tagging systems is often modelled as a graph of connections between users, items and tags. Recommendation algorithms such as FolkRank, have the potential to leverage complex relationships in the data, corresponding to multiple hops in the graph. We present an in-depth analysis and evaluation of graph models for social tagging data and propose novel adaptations and extensions of FolkRank to improve tag recommendations. We highlight implicit assumptions made by the widely used folksonomy model, and propose an alternative and more accurate graph-representation of the data. Our extensions of FolkRank address the new item problem by incorporating content data into the algorithm, and significantly improve prediction results on unpruned datasets. Our adaptations address issues in the iterative weight spreading calculation that potentially hinder FolkRank's ability to leverage the deep graph as an information source. Moreover, we evaluate the benefit of considering each deeper level of the graph, and present important insights regarding the characteristics of social tagging data in general. Our results suggest that the base assumption made by conventional weight propagation methods, that closeness in the graph always implies a positive relationship, does not hold for the social tagging domain.

  @article{landia2013deeper, abstract = {The information contained in social tagging systems is often modelled as a graph of connections between users, items and tags. Recommendation algorithms such as FolkRank, have the potential to leverage complex relationships in the data, corresponding to multiple hops in the graph. We present an in-depth analysis and evaluation of graph models for social tagging data and propose novel adaptations and extensions of FolkRank to improve tag recommendations. We highlight implicit assumptions made by the widely used folksonomy model, and propose an alternative and more accurate graph-representation of the data. Our extensions of FolkRank address the new item problem by incorporating content data into the algorithm, and significantly improve prediction results on unpruned datasets. Our adaptations address issues in the iterative weight spreading calculation that potentially hinder FolkRank's ability to leverage the deep graph as an information source. Moreover, we evaluate the benefit of considering each deeper level of the graph, and present important insights regarding the characteristics of social tagging data in general. Our results suggest that the base assumption made by conventional weight propagation methods, that closeness in the graph always implies a positive relationship, does not hold for the social tagging domain.}, author = {Landia, Nikolas and Doerfel, Stephan and Jäschke, Robert and Anand, Sarabjot Singh and Hotho, Andreas and Griffiths, Nathan}, journal = {cs.IR}, keywords = {tagging}, title = {Deeper Into the Folksonomy Graph: FolkRank Adaptations and Extensions for Improved Tag Recommendations}, volume = {1310.1498}, year = 2013 }
• 
  A Generic Platform for Ubiquitous and Subjective Data M. Becker; J. Mueller; A. Hotho; G. Stumme (2013). New York, NY, USA.
  • [ Abstract ]
  • [ BibTeX ]
  • [ Download ]
  An increasing number of platforms like Xively or ThingSpeak are available to manage ubiquitous sensor data enabling the Internet of Things. Strict data formats allow interoperability and informative visualizations, supporting the development of custom user applications. Yet, these strict data formats as well as the common feed-centric approach limit the flexibility of these platforms. We aim at providing a concept that supports data ranging from text-based formats like JSON to images and video footage. Furthermore, we introduce the concept of extensions, which allows to enrich existing data points with additional information, thus, taking a data point centric approach. This enables us to gain semantic and user specific context by attaching subjective data to objective values. This paper provides an overview of our architecture including concept, implementation details and present applications. We distinguish our approach from several other systems and describe two sensing applications namely AirProbe and WideNoise that were implemented for our platform.

  @inproceedings{mueller-2013a, abstract = {An increasing number of platforms like Xively or ThingSpeak are available to manage ubiquitous sensor data enabling the Internet of Things. Strict data formats allow interoperability and informative visualizations, supporting the development of custom user applications. Yet, these strict data formats as well as the common feed-centric approach limit the flexibility of these platforms. We aim at providing a concept that supports data ranging from text-based formats like JSON to images and video footage. Furthermore, we introduce the concept of extensions, which allows to enrich existing data points with additional information, thus, taking a data point centric approach. This enables us to gain semantic and user specific context by attaching subjective data to objective values. This paper provides an overview of our architecture including concept, implementation details and present applications. We distinguish our approach from several other systems and describe two sensing applications namely AirProbe and WideNoise that were implemented for our platform.}, author = {Becker, Martin and Mueller, Juergen and Hotho, Andreas and Stumme, Gerd}, booktitle = {1st International Workshop on Pervasive Urban Crowdsensing Architecture and Applications, PUCAA 2013, Zurich, Switzerland – September 9, 2013. Proceedings}, keywords = {workshop}, note = {Accepted for publication}, pages = {New York, NY, USA}, publisher = {ACM}, title = {A Generic Platform for Ubiquitous and Subjective Data}, year = 2013 }
• 
  Awareness and Learning in Participatory Noise Sensing M. Becker; S. Caminiti; D. Fiorella; L. Francis; P. Gravino; M. (Muki) Haklay; A. Hotho; V. Loreto; J. Mueller; F. Ricchiuti; V. D. P. Servedio; A. Sîrbu; F. Tria in PLoS ONE (2013). 8(12) e81638.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  <p>The development of ICT infrastructures has facilitated the emergence of new paradigms for looking at society and the environment over the last few years. Participatory environmental sensing, i.e. directly involving citizens in environmental monitoring, is one example, which is hoped to encourage learning and enhance awareness of environmental issues. In this paper, an analysis of the behaviour of individuals involved in noise sensing is presented. Citizens have been involved in noise measuring activities through the WideNoise smartphone application. This application has been designed to record both objective (noise samples) and subjective (opinions, feelings) data. The application has been open to be used freely by anyone and has been widely employed worldwide. In addition, several test cases have been organised in European countries. Based on the information submitted by users, an analysis of emerging awareness and learning is performed. The data show that changes in the way the environment is perceived after repeated usage of the application do appear. Specifically, users learn how to recognise different noise levels they are exposed to. Additionally, the subjective data collected indicate an increased user involvement in time and a categorisation effect between pleasant and less pleasant environments.</p>

  @article{10.1371/journal.pone.0081638, abstract = {<p>The development of ICT infrastructures has facilitated the emergence of new paradigms for looking at society and the environment over the last few years. Participatory environmental sensing, i.e. directly involving citizens in environmental monitoring, is one example, which is hoped to encourage learning and enhance awareness of environmental issues. In this paper, an analysis of the behaviour of individuals involved in noise sensing is presented. Citizens have been involved in noise measuring activities through the WideNoise smartphone application. This application has been designed to record both objective (noise samples) and subjective (opinions, feelings) data. The application has been open to be used freely by anyone and has been widely employed worldwide. In addition, several test cases have been organised in European countries. Based on the information submitted by users, an analysis of emerging awareness and learning is performed. The data show that changes in the way the environment is perceived after repeated usage of the application do appear. Specifically, users learn how to recognise different noise levels they are exposed to. Additionally, the subjective data collected indicate an increased user involvement in time and a categorisation effect between pleasant and less pleasant environments.</p>}, author = {Becker, Martin and Caminiti, Saverio and Fiorella, Donato and Francis, Louise and Gravino, Pietro and Haklay, Mordechai (Muki) and Hotho, Andreas and Loreto, Vittorio and Mueller, Juergen and Ricchiuti, Ferdinando and Servedio, Vito D. P. and Sîrbu, Alina and Tria, Francesca}, journal = {PLoS ONE}, keywords = {widenoise}, month = 12, number = 12, pages = {e81638}, publisher = {Public Library of Science}, title = {Awareness and Learning in Participatory Noise Sensing}, volume = 8, year = 2013 }
• 
  Tag Recommendations for SensorFolkSonomies J. Mueller; S. Doerfel; M. Becker; A. Hotho; G. Stumme (2013). (Vol. 1066)
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  With the rising popularity of smart mobile devices, sensor data-based applications have become more and more popular. Their users record data during their daily routine or specifically for certain events. The application WideNoise Plus allows users to record sound samples and to annotate them with perceptions and tags. The app is being used to document and map the soundscape all over the world. The procedure of recording, including the assignment of tags, has to be as easy-to-use as possible. We therefore discuss the application of tag recommender algorithms in this particular scenario. We show, that this task is fundamentally different from the well-known tag recommendation problem in folksonomies as users do no longer tag fix resources but rather sensory data and impressions. The scenario requires efficient recommender algorithms that are able to run on the mobile device, since Internet connectivity cannot be assumed to be available. Therefore, we evaluate the performance of several tag recommendation algorithms and discuss their applicability in the mobile sensing use-case.

  @inproceedings{mueller2013recommendations, abstract = {With the rising popularity of smart mobile devices, sensor data-based applications have become more and more popular. Their users record data during their daily routine or specifically for certain events. The application WideNoise Plus allows users to record sound samples and to annotate them with perceptions and tags. The app is being used to document and map the soundscape all over the world. The procedure of recording, including the assignment of tags, has to be as easy-to-use as possible. We therefore discuss the application of tag recommender algorithms in this particular scenario. We show, that this task is fundamentally different from the well-known tag recommendation problem in folksonomies as users do no longer tag fix resources but rather sensory data and impressions. The scenario requires efficient recommender algorithms that are able to run on the mobile device, since Internet connectivity cannot be assumed to be available. Therefore, we evaluate the performance of several tag recommendation algorithms and discuss their applicability in the mobile sensing use-case.}, address = {Aachen, Germany}, author = {Mueller, Juergen and Doerfel, Stephan and Becker, Martin and Hotho, Andreas and Stumme, Gerd}, booktitle = {Recommender Systems and the Social Web Workshop at 7th ACM Conference on Recommender Systems, RecSys 2013, Hong Kong, China – October 12-16, 2013. Proceedings}, keywords = {tag}, publisher = {CEUR-WS}, title = {Tag Recommendations for SensorFolkSonomies}, volume = 1066, year = 2013 }
• 
  {Semantics of User Interaction in Social Media} F. Mitzlaff; M. Atzmueller; G. Stumme; A. Hotho in Complex Networks IV, G. Ghoshal, J. Poncela-Casasnovas, R. Tolksdorf (Eds.) (2013). (Vol. 476)
  • [ BibTeX ]
  • [ DOI ]
  @incollection{MASH:13, address = {Heidelberg, Germany}, author = {Mitzlaff, Folke and Atzmueller, Martin and Stumme, Gerd and Hotho, Andreas}, booktitle = {Complex Networks IV}, editor = {Ghoshal, Gourab and Poncela-Casasnovas, Julia and Tolksdorf, Robert}, keywords = {user}, publisher = {Springer Verlag}, series = {Studies in Computational Intelligence}, title = {{Semantics of User Interaction in Social Media}}, volume = 476, year = 2013 }
• 
  Proceedings of the Fifth ACM RecSys Workshop on Recommender Systems and the Social Web co-located with the 7th ACM Conference on Recommender Systems (RecSys 2013), Hong Kong, China, October 13, 2013. B. Mobasher; D. Jannach; W. Geyer; J. Freyne; A. Hotho; S. S. Anand; I. Guy in CEUR Workshop Proceedings (2013). (Vol. 1066) CEUR-WS.org.
  • [ BibTeX ]
  • [ URL ]
  @proceedings{conf/recsys/2013rsweb, booktitle = {RSWeb@RecSys}, editor = {Mobasher, Bamshad and Jannach, Dietmar and Geyer, Werner and Freyne, Jill and Hotho, Andreas and Anand, Sarabjot Singh and Guy, Ido}, keywords = {workshop}, publisher = {CEUR-WS.org}, series = {CEUR Workshop Proceedings}, title = {Proceedings of the Fifth ACM RecSys Workshop on Recommender Systems and the Social Web co-located with the 7th ACM Conference on Recommender Systems (RecSys 2013), Hong Kong, China, October 13, 2013.}, volume = 1066, year = 2013 }
• 
  Computing Semantic Relatedness from Human Navigational Paths: A Case Study on Wikipedia P. Singer; T. Niebler; M. Strohmaier; A. Hotho in International Journal on Semantic Web and Information Systems (IJSWIS) (2013). 9(4) 41–70.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  • [ Download ]
  In this article, the authors present a novel approach for computing semantic relatedness and conduct a large-scale study of it on Wikipedia. Unlike existing semantic analysis methods that utilize Wikipedia’s content or link structure, the authors propose to use human navigational paths on Wikipedia for this task. The authors obtain 1.8 million human navigational paths from a semi-controlled navigation experiment – a Wikipedia-based navigation game, in which users are required to find short paths between two articles in a given Wikipedia article network. The authors’ results are intriguing: They suggest that (i) semantic relatedness computed from human navigational paths may be more precise than semantic relatedness computed from Wikipedia’s plain link structure alone and (ii) that not all navigational paths are equally useful. Intelligent selection based on path characteristics can improve accuracy. The authors’ work makes an argument for expanding the existing arsenal of data sources for calculating semantic relatedness and to consider the utility of human navigational paths for this task.

  @article{singer2013computing, abstract = {In this article, the authors present a novel approach for computing semantic relatedness and conduct a large-scale study of it on Wikipedia. Unlike existing semantic analysis methods that utilize Wikipedia’s content or link structure, the authors propose to use human navigational paths on Wikipedia for this task. The authors obtain 1.8 million human navigational paths from a semi-controlled navigation experiment – a Wikipedia-based navigation game, in which users are required to find short paths between two articles in a given Wikipedia article network. The authors’ results are intriguing: They suggest that (i) semantic relatedness computed from human navigational paths may be more precise than semantic relatedness computed from Wikipedia’s plain link structure alone and (ii) that not all navigational paths are equally useful. Intelligent selection based on path characteristics can improve accuracy. The authors’ work makes an argument for expanding the existing arsenal of data sources for calculating semantic relatedness and to consider the utility of human navigational paths for this task.}, author = {Singer, Philipp and Niebler, Thomas and Strohmaier, Markus and Hotho, Andreas}, journal = {International Journal on Semantic Web and Information Systems (IJSWIS)}, keywords = {semantic}, number = 4, pages = {41–70}, publisher = {IGI Global}, title = {Computing Semantic Relatedness from Human Navigational Paths: A Case Study on Wikipedia}, volume = 9, year = 2013 }
• 
  Ubiquitous Social Media Analysis Third International Workshops, MUSE 2012, Bristol, UK, September 24, 2012, and MSM 2012, Milwaukee, WI, USA, June 25, 2012, Revised Selected Papers M. Atzmueller; A. Chin; D. Helic; A. Hotho (2013). Imprint: Springer, Berlin, Heidelberg.
  • [ BibTeX ]
  • [ URL ]
  @book{atzmueller2013ubiquitous, address = {Berlin, Heidelberg}, editor = {Atzmueller, Martin and Chin, Alvin and Helic, Denis and Hotho, Andreas}, keywords = {workshop}, publisher = {Imprint: Springer}, title = {Ubiquitous Social Media Analysis Third International Workshops, MUSE 2012, Bristol, UK, September 24, 2012, and MSM 2012, Milwaukee, WI, USA, June 25, 2012, Revised Selected Papers}, year = 2013 }
• 
  Exploiting Structural Consistencies with Stacked Conditional Random Fields P. Kluegl; M. Toepfer; F. Lemmerich; A. Hotho; F. Puppe in Mathematical Methodologies in Pattern Recognition and Machine Learning Springer Proceedings in Mathematics & Statistics (2013). 30 111–125.
  • [ Abstract ]
  • [ BibTeX ]
  Conditional Random Fields (CRF) are popular methods for labeling unstructured or textual data. Like many machine learning approaches, these undirected graphical models assume the instances to be independently distributed. However, in real-world applications data is grouped in a natural way, e.g., by its creation context. The instances in each group often share additional structural consistencies. This paper proposes a domain-independent method for exploiting these consistencies by combining two CRFs in a stacked learning framework. We apply rule learning collectively on the predictions of an initial CRF for one context to acquire descriptions of its specific properties. Then, we utilize these descriptions as dynamic and high quality features in an additional (stacked) CRF. The presented approach is evaluated with a real-world dataset for the segmentation of references and achieves a significant reduction of the labeling error.

  @article{kluegl2013exploiting, abstract = {Conditional Random Fields (CRF) are popular methods for labeling unstructured or textual data. Like many machine learning approaches, these undirected graphical models assume the instances to be independently distributed. However, in real-world applications data is grouped in a natural way, e.g., by its creation context. The instances in each group often share additional structural consistencies. This paper proposes a domain-independent method for exploiting these consistencies by combining two CRFs in a stacked learning framework. We apply rule learning collectively on the predictions of an initial CRF for one context to acquire descriptions of its specific properties. Then, we utilize these descriptions as dynamic and high quality features in an additional (stacked) CRF. The presented approach is evaluated with a real-world dataset for the segmentation of references and achieves a significant reduction of the labeling error.}, author = {Kluegl, Peter and Toepfer, Martin and Lemmerich, Florian and Hotho, Andreas and Puppe, Frank}, journal = {Mathematical Methodologies in Pattern Recognition and Machine Learning Springer Proceedings in Mathematics & Statistics}, keywords = {references}, pages = {111-125}, title = {Exploiting Structural Consistencies with Stacked Conditional Random Fields}, volume = 30, year = 2013 }

• 
  The challenge of recommender systems challenges. A. Said; D. Tikk; A. Hotho P. Cunningham, N. J. Hurley, I. Guy, S. S. Anand (Eds.) (2012). 9–10.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{conf/recsys/SaidTH12, author = {Said, Alan and Tikk, Domonkos and Hotho, Andreas}, booktitle = {RecSys}, crossref = {conf/recsys/2012}, editor = {Cunningham, Padraig and Hurley, Neil J. and Guy, Ido and Anand, Sarabjot Singh}, keywords = {sys:relevantfor:l3s}, pages = {9-10}, publisher = {ACM}, title = {The challenge of recommender systems challenges.}, year = 2012 }
• 
  4th ACM RecSys workshop on recommender systems and the social web. B. Mobasher; D. Jannach; W. Geyer; A. Hotho P. Cunningham, N. J. Hurley, I. Guy, S. S. Anand (Eds.) (2012). 345–346.
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{conf/recsys/MobasherJGH12, author = {Mobasher, Bamshad and Jannach, Dietmar and Geyer, Werner and Hotho, Andreas}, booktitle = {RecSys}, crossref = {conf/recsys/2012}, editor = {Cunningham, Padraig and Hurley, Neil J. and Guy, Ido and Anand, Sarabjot Singh}, keywords = {workshop}, pages = {345-346}, publisher = {ACM}, title = {4th ACM RecSys workshop on recommender systems and the social web.}, year = 2012 }
• 
  RSWeb ’12: Proceedings of the 4th ACM RecSys workshop on Recommender systems and the social web B. Mobasher; D. Jannach; W. Geyer; A. Hotho (2012). ACM, Dublin, Ireland.
  • [ Abstract ]
  • [ BibTeX ]
  The new opportunities for applying recommendation techniques within Social Web platforms and applications as well as the various new sources of information which have become available in the Web 2.0 and can be incorporated in future recommender applications are a strong driving factor in current recommender system research for various reasons:</p> <p>(1) Social systems by their definition encourage interaction between users and both online content and other users, thus generating new sources of knowledge for recommender systems. Web 2.0 users explicitly provide personal information and implicitly express preferences through their interactions with others and the system (e.g. commenting, friending, rating, etc.). These various new sources of knowledge can be leveraged to improve recommendation techniques and develop new strategies which focus on social recommendation.</p> <p>(2) New application areas for recommender systems emerge with the popularity of the Social Web. Recommenders cannot only be used to sort and filter Web 2.0 and social network information, they can also support users in the information sharing process, e.g., by recommending suitable tags during folksonomy development.</p> <p>(3) Recommender technology can assist Social Web systems through increasing adoption and participation and sustaining membership. Through targeted and timely intervention which stimulates traffic and interaction, recommender technology can play its role in sustaining the success of the Social Web.</p> <p>(4) The Social Web also presents new challenges for recommender systems, such as the complicated nature of human-to-human interaction which comes into play when recommending people and can require more interactive and richer recommender systems user interfaces.</p> <p>The technical papers appearing in these proceedings aim to explore and understand challenges and new opportunities for recommender systems in the Social Web and were selected in a formal review process by an international program committee.</p> <p>Overall, we received 13 paper submissions from 12 different countries, out of which 7 long papers and 1 short paper were selected for presentation and inclusion in the proceedings. The submitted papers addressed a variety of topics related to Social Web recommender systems from the use of microblogging data for personalization over new tag recommendation approaches to social media-based personalization of news.

  @proceedings{Mobasher:2012:2365934, abstract = {The new opportunities for applying recommendation techniques within Social Web platforms and applications as well as the various new sources of information which have become available in the Web 2.0 and can be incorporated in future recommender applications are a strong driving factor in current recommender system research for various reasons:</p> <p>(1) Social systems by their definition encourage interaction between users and both online content and other users, thus generating new sources of knowledge for recommender systems. Web 2.0 users explicitly provide personal information and implicitly express preferences through their interactions with others and the system (e.g. commenting, friending, rating, etc.). These various new sources of knowledge can be leveraged to improve recommendation techniques and develop new strategies which focus on social recommendation.</p> <p>(2) New application areas for recommender systems emerge with the popularity of the Social Web. Recommenders cannot only be used to sort and filter Web 2.0 and social network information, they can also support users in the information sharing process, e.g., by recommending suitable tags during folksonomy development.</p> <p>(3) Recommender technology can assist Social Web systems through increasing adoption and participation and sustaining membership. Through targeted and timely intervention which stimulates traffic and interaction, recommender technology can play its role in sustaining the success of the Social Web.</p> <p>(4) The Social Web also presents new challenges for recommender systems, such as the complicated nature of human-to-human interaction which comes into play when recommending people and can require more interactive and richer recommender systems user interfaces.</p> <p>The technical papers appearing in these proceedings aim to explore and understand challenges and new opportunities for recommender systems in the Social Web and were selected in a formal review process by an international program committee.</p> <p>Overall, we received 13 paper submissions from 12 different countries, out of which 7 long papers and 1 short paper were selected for presentation and inclusion in the proceedings. The submitted papers addressed a variety of topics related to Social Web recommender systems from the use of microblogging data for personalization over new tag recommendation approaches to social media-based personalization of news.}, address = {New York, NY, USA}, author = {Mobasher, Bamshad and Jannach, Dietmar and Geyer, Werner and Hotho, Andreas}, keywords = {workshop}, note = 609126, publisher = {ACM}, title = {RSWeb '12: Proceedings of the 4th ACM RecSys workshop on Recommender systems and the social web}, year = 2012 }
• 
  Extending FolkRank with content data N. Landia; S. S. Anand; A. Hotho; R. J\"{a}schke; S. Doerfel; F. Mitzlaff in RSWeb ’12 (2012). 1–8.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Real-world tagging datasets have a large proportion of new/ untagged documents. Few approaches for recommending tags to a user for a document address this new item problem, concentrating instead on artificially created post-core datasets where it is guaranteed that the user as well as the document of each test post is known to the system and already has some tags assigned to it. In order to recommend tags for new documents, approaches are required which model documents not only based on the tags assigned to them in the past (if any), but also the content. In this paper we present a novel adaptation to the widely recognised FolkRank tag recommendation algorithm by including content data. We adapt the FolkRank graph to use word nodes instead of document nodes, enabling it to recommend tags for new documents based on their textual content. Our adaptations make FolkRank applicable to post-core 1 ie. the full real-world tagging datasets and address the new item problem in tag recommendation. For comparison, we also apply and evaluate the same methodology of including content on a simpler tag recommendation algorithm. This results in a less expensive recommender which suggests a combination of user related and document content related tags.</p> <p>Including content data into FolkRank shows an improvement over plain FolkRank on full tagging datasets. However, we also observe that our simpler content-aware tag recommender outperforms FolkRank with content data. Our results suggest that an optimisation of the weighting method of FolkRank is required to achieve better results.

  @inproceedings{Landia:2012:EFC:2365934.2365936, abstract = {Real-world tagging datasets have a large proportion of new/ untagged documents. Few approaches for recommending tags to a user for a document address this new item problem, concentrating instead on artificially created post-core datasets where it is guaranteed that the user as well as the document of each test post is known to the system and already has some tags assigned to it. In order to recommend tags for new documents, approaches are required which model documents not only based on the tags assigned to them in the past (if any), but also the content. In this paper we present a novel adaptation to the widely recognised FolkRank tag recommendation algorithm by including content data. We adapt the FolkRank graph to use word nodes instead of document nodes, enabling it to recommend tags for new documents based on their textual content. Our adaptations make FolkRank applicable to post-core 1 ie. the full real-world tagging datasets and address the new item problem in tag recommendation. For comparison, we also apply and evaluate the same methodology of including content on a simpler tag recommendation algorithm. This results in a less expensive recommender which suggests a combination of user related and document content related tags.</p> <p>Including content data into FolkRank shows an improvement over plain FolkRank on full tagging datasets. However, we also observe that our simpler content-aware tag recommender outperforms FolkRank with content data. Our results suggest that an optimisation of the weighting method of FolkRank is required to achieve better results.}, address = {New York, NY, USA}, author = {Landia, Nikolas and Anand, Sarabjot Singh and Hotho, Andreas and J\"{a}schke, Robert and Doerfel, Stephan and Mitzlaff, Folke}, booktitle = {Proceedings of the 4th ACM RecSys workshop on Recommender systems and the social web}, keywords = {tagging}, pages = {1–8}, publisher = {ACM}, series = {RSWeb '12}, title = {Extending FolkRank with content data}, year = 2012 }
• 
  Leveraging publication metadata and social data into FolkRank for scientific publication recommendation S. Doerfel; R. Jäschke; A. Hotho; G. Stumme in RSWeb ’12 (2012). 9–16.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  The ever-growing flood of new scientific articles requires novel retrieval mechanisms. One means for mitigating this instance of the information overload phenomenon are collaborative tagging systems, that allow users to select, share and annotate references to publications. These systems employ recommendation algorithms to present to their users personalized lists of interesting and relevant publications. In this paper we analyze different ways to incorporate social data and metadata from collaborative tagging systems into the graph-based ranking algorithm FolkRank to utilize it for recommending scientific articles to users of the social bookmarking system BibSonomy. We compare the results to those of Collaborative Filtering, which has previously been applied for resource recommendation.

  @inproceedings{Doerfel:2012:LPM:2365934.2365937, abstract = {The ever-growing flood of new scientific articles requires novel retrieval mechanisms. One means for mitigating this instance of the information overload phenomenon are collaborative tagging systems, that allow users to select, share and annotate references to publications. These systems employ recommendation algorithms to present to their users personalized lists of interesting and relevant publications. In this paper we analyze different ways to incorporate social data and metadata from collaborative tagging systems into the graph-based ranking algorithm FolkRank to utilize it for recommending scientific articles to users of the social bookmarking system BibSonomy. We compare the results to those of Collaborative Filtering, which has previously been applied for resource recommendation.}, address = {New York, NY, USA}, author = {Doerfel, Stephan and Jäschke, Robert and Hotho, Andreas and Stumme, Gerd}, booktitle = {Proceedings of the 4th ACM RecSys workshop on Recommender systems and the social web}, keywords = {tagging}, pages = {9–16}, publisher = {ACM}, series = {RSWeb '12}, title = {Leveraging publication metadata and social data into FolkRank for scientific publication recommendation}, year = 2012 }
• 
  Publikationen im Web 2.0 A. Hotho in Informatik-Spektrum (2012). 1–5.
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  @article{hotho2012publikationen, address = {Berlin / Heidelberg}, author = {Hotho, Andreas}, journal = {Informatik-Spektrum}, keywords = {web2.0}, pages = {1-5}, publisher = {Springer}, title = {Publikationen im Web 2.0}, year = 2012 }
• 
  Ubicon: Observing Social and Physical Activities M. Atzmueller; M. Becker; S. Doerfel; M. Kibanov; A. Hotho; B.-E. Macek; F. Mitzlaff; J. Mueller; C. Scholz; G. Stumme (2012).
  • [ Abstract ]
  • [ BibTeX ]
  • [ Download ]
  The connection of ubiquitous and social computing is an emerging research area which is combining two prominent areas of computer science. In this paper, we tackle this topic from different angles: We describe data mining methods for ubiquitous and social data, specifically focusing on physical and social activities, and provide exemplary analysis results. Furthermore, we give an overview on the Ubicon platform which provides a framework for the creation and hosting of ubiquitous and social applications for diverse tasks and projects. Ubicon features the collection and analysis of both physical and social activities of users for enabling inter-connected applications in ubiquitous and social contexts. We summarize three real-world systems built on top of Ubicon, and exemplarily discuss the according mining and analysis aspects.

  @inproceedings{mueller-2012, abstract = {The connection of ubiquitous and social computing is an emerging research area which is combining two prominent areas of computer science. In this paper, we tackle this topic from different angles: We describe data mining methods for ubiquitous and social data, specifically focusing on physical and social activities, and provide exemplary analysis results. Furthermore, we give an overview on the Ubicon platform which provides a framework for the creation and hosting of ubiquitous and social applications for diverse tasks and projects. Ubicon features the collection and analysis of both physical and social activities of users for enabling inter-connected applications in ubiquitous and social contexts. We summarize three real-world systems built on top of Ubicon, and exemplarily discuss the according mining and analysis aspects.}, address = {Washington, DC, USA}, author = {Atzmueller, Martin and Becker, Martin and Doerfel, Stephan and Kibanov, Mark and Hotho, Andreas and Macek, Björn-Elmar and Mitzlaff, Folke and Mueller, Juergen and Scholz, Christoph and Stumme, Gerd}, booktitle = {IEEE International Conference on Cyber, Physical and Social Computing, CPSCom 2012, Besançon, France, 20-23 November, 2012}, keywords = {ubicon}, publisher = {IEEE}, title = {Ubicon: Observing Social and Physical Activities}, year = 2012 }
• 
  Recommender Systems for Social Tagging Systems L. Balby Marinho; A. Hotho; R. Jäschke; A. Nanopoulos; S. Rendle; L. Schmidt-Thieme; G. Stumme; P. Symeonidis in SpringerBriefs in Electrical and Computer Engineering (2012). Springer.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  Social Tagging Systems are web applications in which users upload resources (e.g., bookmarks, videos, photos, etc.) and annotate it with a list of freely chosen keywords called tags. This is a grassroots approach to organize a site and help users to find the resources they are interested in. Social tagging systems are open and inherently social; features that have been proven to encourage participation. However, with the large popularity of these systems and the increasing amount of user-contributed content, information overload rapidly becomes an issue. Recommender Systems are well known applications for increasing the level of relevant content over the “noise” that continuously grows as more and more content becomes available online. In social tagging systems, however, we face new challenges. While in classic recommender systems the mode of recommendation is basically the resource, in social tagging systems there are three possible modes of recommendation: users, resources, or tags. Therefore suitable methods that properly exploit the different dimensions of social tagging systems data are needed. In this book, we survey the most recent and state-of-the-art work about a whole new generation of recommender systems built to serve social tagging systems. The book is divided into self-contained chapters covering the background material on social tagging systems and recommender systems to the more advanced techniques like the ones based on tensor factorization and graph-based models.

  @book{balbymarinho2012recommender, abstract = {Social Tagging Systems are web applications in which users upload resources (e.g., bookmarks, videos, photos, etc.) and annotate it with a list of freely chosen keywords called tags. This is a grassroots approach to organize a site and help users to find the resources they are interested in. Social tagging systems are open and inherently social; features that have been proven to encourage participation. However, with the large popularity of these systems and the increasing amount of user-contributed content, information overload rapidly becomes an issue. Recommender Systems are well known applications for increasing the level of relevant content over the “noise” that continuously grows as more and more content becomes available online. In social tagging systems, however, we face new challenges. While in classic recommender systems the mode of recommendation is basically the resource, in social tagging systems there are three possible modes of recommendation: users, resources, or tags. Therefore suitable methods that properly exploit the different dimensions of social tagging systems data are needed. In this book, we survey the most recent and state-of-the-art work about a whole new generation of recommender systems built to serve social tagging systems. The book is divided into self-contained chapters covering the background material on social tagging systems and recommender systems to the more advanced techniques like the ones based on tensor factorization and graph-based models.}, author = {Balby Marinho, L. and Hotho, A. and Jäschke, R. and Nanopoulos, A. and Rendle, S. and Schmidt-Thieme, L. and Stumme, G. and Symeonidis, P.}, keywords = {tagging}, month = {feb}, publisher = {Springer}, series = {SpringerBriefs in Electrical and Computer Engineering}, title = {Recommender Systems for Social Tagging Systems}, year = 2012 }
• 
  Face-to-Face Contacts at a Conference: Dynamics of Communities and Roles M. Atzmueller; S. Doerfel; A. Hotho; F. Mitzlaff; G. Stumme in {Modeling and Mining Ubiquitous Social Media} (2012). (Vol. 7472)
  • [ BibTeX ]
  • [ URL ]
  @incollection{ADHMS:12, address = {Heidelberg, Germany}, author = {Atzmueller, Martin and Doerfel, Stephan and Hotho, Andreas and Mitzlaff, Folke and Stumme, Gerd}, booktitle = {{Modeling and Mining Ubiquitous Social Media}}, keywords = {sys:relevantfor:l3s}, publisher = {Springer Verlag}, series = {LNAI}, title = {Face-to-Face Contacts at a Conference: Dynamics of Communities and Roles}, volume = 7472, year = 2012 }
• 
  Collective Information Extraction with Context-Specific Consistencies. P. Klügl; M. Toepfer; F. Lemmerich; A. Hotho; F. Puppe in Lecture Notes in Computer Science, P. A. Flach, T. D. Bie, N. Cristianini (Eds.) (2012). (Vol. 7523) 728–743.
  • [ BibTeX ]
  • [ URL ]
  • [ Download ]
  @inproceedings{conf/pkdd/KluglTLHP12, author = {Klügl, Peter and Toepfer, Martin and Lemmerich, Florian and Hotho, Andreas and Puppe, Frank}, booktitle = {ECML/PKDD (1)}, crossref = {conf/pkdd/2012-1}, editor = {Flach, Peter A. and Bie, Tijl De and Cristianini, Nello}, keywords = {tau}, pages = {728-743}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, title = {Collective Information Extraction with Context-Specific Consistencies.}, volume = 7523, year = 2012 }
• 
  Datenschutz im Web 2.0 am Beispiel des sozialen Tagging-Systems BibSonomy. B. Krause; H. Lerch; A. Hotho; A. Roßnagel; G. Stumme in Informatik Spektrum (2012). 35(1) 12–23.
  • [ BibTeX ]
  • [ URL ]
  @article{journals/insk/KrauseLHRS12, author = {Krause, Beate and Lerch, Hana and Hotho, Andreas and Roßnagel, Alexander and Stumme, Gerd}, journal = {Informatik Spektrum}, keywords = {tagging}, number = 1, pages = {12-23}, title = {Datenschutz im Web 2.0 am Beispiel des sozialen Tagging-Systems BibSonomy.}, volume = 35, year = 2012 }
• 
  Modeling and Mining Ubiquitous Social Media M. Atzmueller; A. Chin; D. Helic; A. Hotho in Lecture Notes in Computer Science (2012). (Vol. 7472) Springer Verlag, Heidelberg, Germany.
  • [ BibTeX ]
  • [ URL ]
  @book{ACHH:12, address = {Heidelberg, Germany}, editor = {Atzmueller, Martin and Chin, Alvin and Helic, Denis and Hotho, Andreas}, keywords = {ubiquitous}, publisher = {Springer Verlag}, series = {Lecture Notes in Computer Science}, title = {Modeling and Mining Ubiquitous Social Media}, volume = 7472, year = 2012 }
• 
  {Proceedings of the Third International Workshop on Mining Ubiquitous and Social Environments (MUSE 2012)} M. Atzmueller; A. Hotho (2012). Workshop Notes, Bristol, UK.
  • [ BibTeX ]
  • [ URL ]
  @book{AH:12, address = {Bristol, UK}, editor = {Atzmueller, Martin and Hotho, Andreas}, keywords = {sys:relevantfor:l3s}, publisher = {Workshop Notes}, title = {{Proceedings of the Third International Workshop on Mining Ubiquitous and Social Environments (MUSE 2012)}}, year = 2012 }
• 
  Challenges in Tag Recommendations for Collaborative Tagging Systems R. Jäschke; A. Hotho; F. Mitzlaff; G. Stumme in Recommender Systems for the Social Web, J. J. Pazos Arias, A. Fernández Vilas, R. P. Díaz Redondo (Eds.) (2012). (Vol. 32) 65–87.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  • [ DOI ]
  Originally introduced by social bookmarking systems, collaborative tagging, or social tagging, has been widely adopted by many web-based systems like wikis, e-commerce platforms, or social networks. Collaborative tagging systems allow users to annotate resources using freely chosen keywords, so called tags . Those tags help users in finding/retrieving resources, discovering new resources, and navigating through the system. The process of tagging resources is laborious. Therefore, most systems support their users by tag recommender components that recommend tags in a personalized way. The Discovery Challenges 2008 and 2009 of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) tackled the problem of tag recommendations in collaborative tagging systems. Researchers were invited to test their methods in a competition on datasets from the social bookmark and publication sharing system BibSonomy. Moreover, the 2009 challenge included an online task where the recommender systems were integrated into BibSonomy and provided recommendations in real time. In this chapter we review, evaluate and summarize the submissions to the two Discovery Challenges and thus lay the groundwork for continuing research in this area.

  @incollection{jaeschke2012challenges, abstract = {Originally introduced by social bookmarking systems, collaborative tagging, or social tagging, has been widely adopted by many web-based systems like wikis, e-commerce platforms, or social networks. Collaborative tagging systems allow users to annotate resources using freely chosen keywords, so called tags . Those tags help users in finding/retrieving resources, discovering new resources, and navigating through the system. The process of tagging resources is laborious. Therefore, most systems support their users by tag recommender components that recommend tags in a personalized way. The Discovery Challenges 2008 and 2009 of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) tackled the problem of tag recommendations in collaborative tagging systems. Researchers were invited to test their methods in a competition on datasets from the social bookmark and publication sharing system BibSonomy. Moreover, the 2009 challenge included an online task where the recommender systems were integrated into BibSonomy and provided recommendations in real time. In this chapter we review, evaluate and summarize the submissions to the two Discovery Challenges and thus lay the groundwork for continuing research in this area.}, address = {Berlin/Heidelberg}, author = {Jäschke, Robert and Hotho, Andreas and Mitzlaff, Folke and Stumme, Gerd}, booktitle = {Recommender Systems for the Social Web}, editor = {Pazos Arias, José J. and Fernández Vilas, Ana and Díaz Redondo, Rebeca P.}, keywords = {tagging}, pages = {65–87}, publisher = {Springer}, series = {Intelligent Systems Reference Library}, title = {Challenges in Tag Recommendations for Collaborative Tagging Systems}, volume = 32, year = 2012 }
• 
  Stacked Conditional Random Fields Exploiting Structural Consistencies P. Klügl; M. Toepfer; F. Lemmerich; A. Hotho; F. Puppe P. L. Carmona, J. S. Sánchez, A. Fred (Eds.) (2012). 240–248.
  • [ Abstract ]
  • [ BibTeX ]
  • [ URL ]
  Conditional Random Fields CRF are popular methods for labeling unstructured or textual data. Like many machine learning approaches these undirected graphical models assume the instances to be independently distributed. However, in real world applications data is grouped in a natural way, e.g., by its creation context. The instances in each group often share additional structural consistencies. This paper proposes a domain-independent method for exploiting these consistencies by combining two CRFs in a stacked learning framework. The approach incorporates three successive steps of inference: First, an initial CRF processes single instances as usual. Next, we apply rule learning collectively on all labeled outputs of one context to acquire descriptions of its specific properties. Finally, we utilize these descriptions as dynamic and high quality features in an additional stacked CRF. The presented approach is evaluated with a real-world dataset for the segmentation of references and achieves a significant reduction of the labeling error.

  @inproceedings{kluegl2012stacked, abstract = {Conditional Random Fields CRF are popular methods for labeling unstructured or textual data. Like many machine learning approaches these undirected graphical models assume the instances to be independently distributed. However, in real world applications data is grouped in a natural way, e.g., by its creation context. The instances in each group often share additional structural consistencies. This paper proposes a domain-independent method for exploiting these consistencies by combining two CRFs in a stacked learning framework. The approach incorporates three successive steps of inference: First, an initial CRF processes single instances as usual. Next, we apply rule learning collectively on all labeled outputs of one context to acquire descriptions of its specific properties. Finally, we utilize these descriptions as dynamic and high quality features in an additional stacked CRF. The presented approach is evaluated with a real-world dataset for the segmentation of references and achieves a significant reduction of the labeling error.}, address = {Vilamoura, Algarve, Portugal}, author = {Klügl, Peter and Toepfer, Martin and Lemmerich, Florian and Hotho, Andreas and Puppe, Frank}, booktitle = {Proceedings of 1st International Conference on Pattern Recognition Applications and Methods ICPRAM}, editor = {Carmona, Pedro Latorre and Sánchez, J. Salvador and Fred, Ana}, keywords = {sys:relevantfor:uniwue_info6}, pages = {240-248}, publisher = {SciTePress}, title = {Stacked Conditional Random Fields Exploiting Structural Consistencies}, year = 2012 }

• 
  Towards Mining Semantic Maturity in Social Bookmarking Systems M. Atzmueller; D. Benz; A. Hotho; G. Stumme A. Passant, S. Fernández, J. Breslin, U. Bojārs (Eds.) (2011).
  • [ BibTeX ]
  • [ URL ]
  @inproceedings{atzmueller2011towards, author = {Atzmueller, Martin and Benz, Dominik and Hotho, Andreas and Stumme, Gerd}, booktitle = {Proceedings of the 4th international workshop on Social Data on the Web (SDoW2011)}, editor = {Passant, Alexandre and Fernández, Sergio and Breslin, John and Bojārs, Uldis}, keyw