1.
Seufert, M., Kargl, J., Schauer, J., Nüchter, A., Hoßfeld, T.: Different Points of View: Impact of 3D Point Cloud Reduction on QoE of Rendered Images. 12th International Conference on Quality of Multimedia Experience (QoMEX). , Athlone, Ireland (2020).
Modern photogrammetric methods as well as laser measurement systems make it easy to collect large 3D point clouds that sample objects or environments. As the recorded point clouds can be used to render computer-generated images and models, they are of particular interest in the domains of geographical and architectural engineering, as well for computer graphics (e.g., games or virtual reality). However, point clouds have a huge storage demand, thus, point clouds shall be reduced by removing some of the points. This will inevitably also reduce the Quality of Experience (QoE) of media, which is rendered from the reduced point clouds. In this work, the impact of two different reduction methods on the QoE of rendered images is investigated from two point of views, i.e., based on ratings from both naive crowdworkers as well as point cloud experts.
2.
Wenninger, S., Achenbach, J., Bartl, A., Latoschik, M.E., Botsch, M.: Realistic Virtual Humans from Smartphone Videos. In: Teather, R.J., Joslin, C., Stuerzlinger, W., Figueroa, P., Hu, Y., Batmaz, A.U., Lee, W., und Ortega, F. (hrsg.) VRST. S. 29:1–29:11. ACM (2020).
This paper introduces an automated 3D-reconstruction method for generating high-quality virtual humans from monocular smartphone cameras. The input of our approach are two video clips, one capturing the whole body and the other providing detailed close-ups of head and face. Optical flow analysis and sharpness estimation select individual frames, from which two dense point clouds for the body and head are computed using multi-view reconstruction. Automatically detected landmarks guide the fitting of a virtual human body template to these point clouds, thereby reconstructing the geometry. A graph-cut stitching approach reconstructs a detailed texture. Our results are compared to existing low-cost monocular approaches as well as to expensive multi-camera scan rigs. We achieve visually convincing reconstructions that are almost on par with complex camera rigs while surpassing similar low-cost approaches. The generated high-quality avatars are ready to be processed, animated, and rendered by standard XR simulation and game engines such as Unreal or Unity
3.
Wehner, N.: Scoring High: Analysis and Prediction of Viewer Behavior and Engagement in the Context of 2018 FIFA WC Live Streaming. , Proceedings of the 28th ACM International Conference on Multimedia (MM), Seattle, WA, USA (2020).
Large-scale events pose severe challenges to live video streaming service providers, who need to cope with high, peaking viewer numbers and the resulting fluctuating resource demands, keeping high levels of Quality of Experience (QoE) to avoid end-user frustration and churn. In this paper, we analyze a unique dataset consisting of more than a million 2018 FIFA World Cup mobile live streaming sessions, collected at a large national public broadcaster. Different from previous work, we analyze QoE and user engagement as well as their interaction, in dependency to specific soccer match events, which have the potential to trigger flash crowds during a match. Flash crowds are a particular challenge to video service providers, since they cause sudden load peaks and consequently, the likelihood of quality problems. We further exploit the data to model viewer engagement over the course of a soccer match, and show that client counts follow very similar patterns of change across all matches. We believe that the analysis as well as the resulting models are valuable sources of insight for service providers, equipping them with tools for customer-centric resource and capacity management.
4.
Wehner, N.: Inferring Web QoE with Machine Learning from Encrypted Network Traffic. , 2nd KuVS Workshop on Machine Learning for Networking, Würzburg, Germany (2020).
In this work, the problem of Quality of Experience (QoE) monitoring of web browsing is addressed. In particular, the inference of common Web QoE metrics like RUMSI with machine learning is investigated. Based on a large dataset collected with WebPageTest on three different devices, a unique feature set is used to approximate Web QoE metrics with regression and classification approaches. This work highlights work in progress.
5.
Wehner, N., Seufert, M., Schüler, J., Wassermann, S., Casas, P., Hoßfeld, T.: Improving Web QoE Monitoring for Encrypted Network Traffic through Time Series Modeling. ACM SIGMETRICS Performance Evaluation Review. (2020).
This paper addresses the problem of Quality of Experience (QoE) monitoring for web browsing. In particular, the inference of common Web QoE metrics such as Speed Index (SI) is investigated. Based on a large dataset collected with open web-measurement platforms on different device-types, a unique feature set is designed and used to estimate the RUMSI -- an efficient approximation to SI, with machine-learning based regression and classification approaches. Results indicate that it is possible to estimate the RUMSI accurately, and that in particular, recurrent neural networks are highly suitable for the task, as they capture the network dynamics more precisely.
6.
Vomhoff, V.: Reliability of Crowdsourced Measurements, (2020).
7.
Supervisor: Loh, F.: Solving the Collision Problem with Aggregation? A Measurement Driven LoRaWAN Study., (2020).
8.
Brefeld, U., Fromont, Élisa, Hotho, A., Knobbe, A.J., Maathuis, M.H., Robardet, C. hrsg.: Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part III. Springer (2020).
9.
Vetter, A., Götz, D., von Mammen, S.: The Art of Loci: Immerse to Memorize. 2020 IEEE Conference on Games (CoG). S. 642–645. IEEE, Osaka, Japan (2020).
There are uncountable things to be remembered, but most people were never shown how to memorize effectively. With this paper, the application The Art of Loci (TAoL) is presented, to provide the means for efficient and fun learning. In a virtual reality (VR) environment users can express their study material visually, either creating dimensional mind maps or experimenting with mnemonic strategies, such as mind palaces. We also present common memorization techniques in light of their underlying pedagogical foundations and discuss the respective features of TAoL in comparison with similar software applications.
10.
Kunz, F., Hirth, M., Schweitzer, T., Linz, C., Goetz, B., Stellzig-Eisenhauer, A., Borchert, K., Böhm, H.: Subjective perception of craniofacial growth asymmetries in patients with deformational plagiocephaly. Springer Clinical Oral Investigations. 24, https://doi.org/10.1007/s00784–020 (2020).
11.
Stulier, N.: Characterisation of the APM in videogames using DOTA 2, (2020).
12.
Stauffert, J.-P., Niebling, F., Latoschik, M.E.: Latency and Cybersickness: Impact, Causes, and Measures. A Review. Frontiers in Virtual Reality. 1, 31 (2020).
Latency is a key characteristic inherent to any computer system. Motion-to-Photon (MTP) latency describes the time between the movement of a tracked object and its corresponding movement rendered and depicted by computer-generated images on a graphical output screen. High MTP latency can cause a loss of performance in interactive graphics applications and, even worse, can provoke cybersickness in Virtual Reality (VR) applications. Here, cybersickness can degrade VR experiences or may render the experiences completely unusable. It can confound research findings of an otherwise sound experiment. Latency as a contributing factor to cybersickness needs to be properly understood. Its effects need to be analyzed, its sources need to be identified, good measurement methods need to be developed, and proper counter measures need to be developed in order to reduce potentially harmful impacts of latency on the usability and safety of VR systems. Research shows that latency can exhibit intricate timing patterns with various spiking and periodic behavior. These timing behaviors may vary, yet most are found to provoke cybersickness. Overall, latency can differ drastically between different systems interfering with generalization of measurement results. This review article describes the causes and effects of latency with regard to cybersickness. We report on different existing approaches to measure and report latency. Hence, the article provides readers with the knowledge to understand and report latency for their own applications, evaluations, and experiments. It should also help to measure, identify, and finally control and counteract latency and hence gain confidence into the soundness of empirical data collected by VR exposures. Low latency increases the usability and safety of VR systems.
13.
Kobs, K., Koopmann, T., Zehe, A., Fernes, D., Krop, P., Hotho, A.: Where to Submit? Helping Researchers to Choose the Right Venue. Findings of the Association for Computational Linguistics: EMNLP 2020. S. 878–883. Association for Computational Linguistics, Online (2020).
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.
14.
Kobs, K., Koopmann, T., Zehe, A., Fernes, D., Krop, P., Hotho, A.: Where to Submit? Helping Researchers to Choose the Right Venue. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: Findings. S. 878–883. Association for Computational Linguistics, Online (2020).
15.
Steinhaeusser, S.C., Lugrin, B.: Horror Laboratory and Forest Cabin - A Horror Game Series for Desktop Computer, Virtual Reality, and Smart Substitutional Reality. Extended Abstracts of the 2020 Annual Symposium on Computer-Human Interaction in Play. ACM (2020).
16.
Wassermann, S., Seufert, M., Casas, P., Li, G., Kuang, L.: ViCrypt to the Rescue: Real-time, Machine-Learning-driven Video-QoE Monitoring for Encrypted Streaming Traffic. IEEE Transactions on Network and Service Management. 17, 2007–2023 (2020).
Video streaming is the killer application of the Internet today. In this paper, we address the problem of real-time, passive Quality-of-Experience (QoE) monitoring of HTTP Adaptive Video Streaming (HAS), from the Internet-Service-Provider (ISP) perspective - i.e., relying exclusively on in-network traffic measurements. Given the wide adoption of end-to-end encryption, we resort to machine-learning (ML) models to estimate multiple key video-QoE indicators (KQIs) from the analysis of the encrypted traffic. We present ViCrypt, an ML-driven monitoring solution able to infer the most important KQIs for HTTP Adaptive Streaming (HAS), namely stalling, initial delay, video resolution, and average video bitrate. ViCrypt performs estimations in real-time, during the playback of an ongoing video-streaming session, with a fine-grained temporal resolution of just one second. For this, it relies on lightweight, stream-like features continuously extracted from the encrypted stream of packets. Empirical evaluations on a large and heterogeneous corpus of YouTube measurements show that ViCrypt can infer the targeted KQIs with high accuracy, enabling large-scale passive video-QoE monitoring and proactive QoE-aware traffic management. Different from the state of the art, and besides real-time operation, ViCrypt is not bound to coarse-grained KQI-classes, providing better and sharper insights than other solutions. Finally, ViCrypt does not require chunk-detection approaches for feature extraction, significantly reducing the complexity of the monitoring approach, and potentially improving on generalization to different HAS protocols used by other video-streaming services such as Netflix and Amazon.
17.
Wehner, N., Seufert, M., Schüler, J., Wassermann, S., Casas, P., Hoßfeld, T.: Improving Web QoE Monitoring for Encrypted Network Traffic through Time Series Modeling. 2nd Workshop on AI in Networks and Distributed Systems (WAIN). , Milan, Italy (2020).
This paper addresses the problem of Quality of Experience (QoE) monitoring for web browsing. In particular, the inference of common Web QoE metrics such as Speed Index (SI) is investigated. Based on a large dataset collected with open web-measurement platforms on different device-types, a unique feature set is designed and used to estimate the RUMSI -- an efficient approximation to SI, with machine-learning based regression and classification approaches. Results indicate that it is possible to estimate the RUMSI accurately, and that in particular, recurrent neural networks are highly suitable for the task, as they capture the network dynamics more precisely.
18.
Khelloqi, S.: Data Augmentation for Machine Learning on Network Traffic, (2020).
Contact/Kontakt: michael.seufert@uni-wuerzburg.de ### Title: Data Augmentation for Machine Learning on Network Traffic (Deutsche Beschreibung weiter unten) Abstract: Machine learning models need huge amounts of data for the training of models to succeed. One approach to increasing the amount of training data is to take already available data and alter them slightly without changing the label of the data. This so called “data augmentation” is especially popular for image data. For example, an existing image of a bird can be slightly rotated or resized to create a new training example (alteration of available data), which has the same label (bird). In this thesis, similar methods for data augmentation of network traffic shall be developed and implemented. ### Titel: Datenvermehrung für maschinelles Lernen auf Netzverkehr Inhalt: Maschinelles Lernen benötigt sehr große Datenmenge, damit Modelle erfolgreich trainiert werden können. Ein Ansatz um die Menge an Trainingsdaten zu erhöhen ist es, bereits verfügbare Daten zu nehmen und sie leicht zu modifizieren ohne das Label, d.h., die Klasse der Daten, zu verändern. Diese sogenannte Datenvermehrung oder „Data Augmentation“ ist besonders bei Bilddaten beliebt. Zum Beispiel kann ein existierendes Bild eines Vogels leicht gedreht oder skaliert werden, um ein neues Trainingsbeispiel zu erhalten (Veränderung von existierenden Daten), das die selbe Klasse (Vogel) besitzt. In dieser Arbeit sollen ähnliche Methoden für die Datenvermehrung von Netzverkehr entwickelt und implementiert werden.
19.
Supervisor: Dietz, K.: Aufsetzen eines Tcpreplay-Servers, (2020).
Tcpreplay [1, 2] ist ein Tool zur erneuten Wiedergabe und zum Editieren von sogenannten "packet captures" (pcaps), also aufgezeichnetem Netzverkehr. Ursprünglich war Tcpreplay im Bereich der Network Security für das wiederholte Abspielen auffälliger Verkehrsmuster gedacht [2], um Algorithmen zur Erkennung von solchen Netzangriffen zu trainieren. Um derartige Use-Cases besser zu verstehen und zu untersuchen, ist das Ziel des Praktikums zunächst, einen Tcpreplay Server aufzusetzen, um das Abspielen von pcaps zu ermöglichen. Die Funktionalität des Servers soll danach so erweitert werden, dass ausgewählte Charakteristika der pcaps verändert werden können, ohne das resultierende Applikationsverhalten eines verbundenen Clients zu verändern. Anschließend soll eine Validierung der Funktionalität der Wiedergabe im Testbed stattfinden. [1] https://github.com/appneta/tcpreplay [2] https://tcpreplay.appneta.com/
20.
Münster, S., Niebling, F., Bruschke, J., Barthel, K., Friedrichs, K., Kröber, C., Maiwald, F.: Urban History Research and Discovery in the Age of Digital Repositories. A Report About Users and Requirements. In: Kremers, H. (hrsg.) Digital Cultural Heritage. S. 63–84. Springer International Publishing, Cham (2020).
The research group on four-dimensional research and communication of urban history (HistStadt4D) investigates and develops methods and technologies to transfer extensive repositories of historical photographs and their contextual information into a three-dimensional spatial model, with an additional temporal component. This will make content accessible to researchers and the public, via a 4D browser as well as a location-dependent augmented reality representation. Against this background, this article highlights users and requirements of both scholarly and touristic usage of digital information about urban history, in particular historical photographs.
21.
Supervisor: Metzger, F.: Profiling the E2E Delays of Streamed Video Games with the Camera Method, (2020).
Determining the E2E lag of actions in video games is not trivial since there is no built-in way to grasp the full lag, including that contributed by the input and output. A standardized method is to record both input and output with a high framerate camera and count the frames from a visible input event to the result on screen. Your job is to perform these measurements for different actions in a series of games, and examine the overhead game streaming generates for them. Note: It might currently not be possible to perform these experiments, as they rely on a physical testbench.
22.
Supervisor: Metzger, F.: Examining the Netcode of Old, Open-sourced Video Games, (2020).
Video games have a long history of often competitive multiplayer modes, supported by various network implementations in their engines. The source code of many of these games has been open sourced over the years, opening their netcode up for examination. Your job is to extract the netcode from some of these games and compare them to each other. We want to find out, how the employed mechanisms have changed over time and how well they would fare in today's environment.
23.
Supervisor: Metzger, F.: Measuring the Effectiveness of Latency Mitigation Methods with a Prototype Video Game, (2020).
Video games with a competitive multiplayer modes must deal with the troubles of high and varying end-to-end lag. They can employ varying methods to manage, mitigate or conceal this lag and equalize it between different clients. But the closed source nature of video games makes examining the effectiveness of these mechanisms hard. Your job is to implement a simple video game prototype from scratch that uses an authoritative client-server multiplayer approach and implements such lag mitigation mechanisms. With this prototype we want to get a better understanding of the effectiveness of these lag concealment measures.
24.
Gall, D., Preßler, J., Hurtienne, J., Latoschik, M.E.: Self-organizing knowledge management might improve the quality of person-centered dementia care: A qualitative study. International Journal of Medical Informatics. 139, 104132 (2020).
Background: In institutional dementia care, person-centered care improves care processes and the quality of life of residents. However, communication gaps impede the implementation of person-centered care in favor of routinized care. Objective: We evaluated whether self-organizing knowledge management reduces communication gaps and improves the quality of person-centered dementia care. Method: We implemented a self-organizing knowledge management system. Eight significant others of residents with severe dementia and six professional caregivers used a mobile application for six months. We conducted qualitative interviews and focus groups afterward. Main findings: Participants reported that the system increased the quality of person-centered care, reduced communication gaps, increased the task satisfaction of caregivers and the wellbeing of significant others. Conclusions: Based on our findings, we develop the following hypotheses: Self-organizing knowledge management might provide a promising tool to improve the quality of person-centered care. It might reduce communication barriers that impede person-centered care. It might allow transferring content-maintaining tasks from caregivers to significant others. Such distribution of tasks, in turn, might be beneficial for both parties. Furthermore, shared knowledge about situational features might guide person-centered interventions.
25.
Supervisor: Loh, F.: Studying the Live Streaming Behavior of Mobile Twitch.tv and YouTube, (2020).
Mit Hilfe eines vorhandenen Testbeds soll in der Arbeit YouTube und Twitch Live gemessen werden. Basierend auf den Messdaten ist die Frage zu beantworten, was der Unterschied zu On-Demand Streaming ist, wie das Video-Anfrageverhalten ist, sowie was die Unterschiede zwischen Twitch und YouTube Live Streaming sind.
26.
Schüler, J.: Flow Statistics Calculation in a Programmable Data Plane with P4, (2020).
27.
Hofmann, J.: Deep Reinforcement Learning for Configuration of Time-Sensitive-Networking, (2020).
28.
Kobs, K., Zehe, A., Bernstetter, A., Chibane, J., Pfister, J., Tritscher, J., Hotho, A.: Emote-Controlled: Obtaining Implicit Viewer Feedback through Emote based Sentiment Analysis on Comments of Popular Twitch.tv Channels. ACM Transactions on Social Computing. 3, 1–34 (2020).
29.
Furdek, M., Natalino, C., Lipp, F., Hock, D., Di Giglio, A., Schiano, M.: Machine Learning for Optical Network Security Monitoring: A Practical Perspective. Journal of Lightwave Technology. 38, 2860–2871 (2020).
30.
von der Pütten, A.M.R., Lugrin, B., Steinhaeusser, S.C., Klass, L.: Context Matters! Identifying Social Context Factors and Assessing Their Relevance for a Socially Assistive Robot. Companion of the 2020 International Conference on Human-Robot Interaction. S. 409–411. ACM (2020).
31.
Münster, S., Maiwald, F., Lehmann, C., Lazariv, T., Hofmann, M., Niebling, F.: An Automated Pipeline for a Browser-Based, City-Scale Mobile 4D VR Application Based on Historical Images. Proceedings of the 2nd Workshop on Structuring and Understanding of Multimedia HeritAge Contents. S. 33–40. Association for Computing Machinery, Seattle, WA, USA (2020).
The process for automatically creating 3D city models from contemporary photographs and visualizing them on mobile devices is now well established, but historical 4D city models are more challenging. The fourth dimension here is time. This article describes an automated VR pipeline based on historical photographs and resulting in an interactive browser-based device-rendered 4D visualization and information system for mobile devices. Since the pipeline shown is currently still under development, initial results for stages of the process will be shown and assessed for accuracy and usability.
32.
Hoßfeld, T., Heegaard, P.E., Skorin-Kapov, L., Varela, M.: Deriving QoE in systems: from fundamental relationships to a QoE‑based Service‑level Quality Index. Springer Quality and User Experience. 5, https://rdcu.be/b483m (2020).
With Quality of Experience (QoE) research having made significant advances over the years, service and network providers aim at user-centric evaluation of the services provided in their system. The question arises how to derive QoE in systems. In the context of subjective user studies conducted to derive relationships between influence factors and QoE, user diversity leads to varying distributions of user rating scores for different test conditions. Such models are commonly exploited by providers to derive various QoE metrics in their system, such as expected QoE, or the percentage of users rating above a certain threshold. The question then becomes how to combine (a) user rating distributions obtained from subjective studies, and (b) system parameter distributions, so as to obtain the actual observed QoE distribution in the system? Moreover, how can various QoE metrics of interest in the system be derived? We prove fundamental relationships for the derivation of QoE in systems, thus providing an important link between the QoE community and the systems community. In our numerical examples, we focus mainly on QoE metrics. We furthermore provide a more generalized view on quantifying the quality of systems by defining a QoE-based Service-level Quality Index. This index exploits the fact that quality can be seen as a proxy measure for utility. Following the assumption that not all user sessions should be weighted equally, we aim to provide a generic framework that can be utilized to quantify the overall utility of a service delivered by a system.
33.
Stauffert, J.-P., Niebling, F., Latoschik, M.E.: Simultaneous Run-Time Measurement of Motion-to-Photon Latency and Latency Jitter. 2020 IEEE Conference on Virtual Reality and 3D User Interfaces (VR). S. 636–644. IEEE (2020).
34.
Stauffert, J.-P., Niebling, F., Lugrin, J.-L., Latoschik, M.E.: Guided Sine Fitting for Latency Estimation in Virtual Reality. 2020 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW). S. 707–708. IEEE (2020).
35.
Janiak, L.: Potential Traffic Savings by Leveraging Proximity of Communication Groups in Mobile Messaging Applications, (2020).
36.
Schwind, A., Wamser, F., Hoßfeld, T., Wunderer, S., Tarnvik, E., Hall, A.: Crowdsourced Network Measurements in Germany: Mobile Internet Experience from End User Perspective. Broadband Coverage in Germany; 14. ITG Symposium. , Berlin (2020).
Collecting and analyzing meaningful data in mobile networks is the key to assessing network performance. Crowdsourced Network Measurements (CNMs) provide insights beyond the network layer and offer performance and other measurements at the application and user-level towards Quality of Experience (QoE). In this paper, the mobile Internet experience for Germany is evaluated with the help of crowdsourcing from the perspective of an end user. We statistically analyze a dataset with throughput measurements on the end device from Tutela Ltd., which covers more than 2.5 million throughput tests across Germany from January to July 2019. We give insights into this emerging methodology and highlight the benefits of this method. The paper contains statistics and conclusions for several large cities as well as regions in Germany compared to general statements for Germany, since individual measurements and averages often only imprecisely reflect the situation. The goal is to give a holistic view of the performance of the current mobile network in Germany. Reading this paper, it becomes evident that reliable statements about the quality of the mobile network for Germany depend on a large number of peculiarities in different regions with their own performance characteristics due to different network deployments and population numbers.
37.
Beyer, S., Chimani, M., Spoerhase, J.: A Simple Primal-Dual Approximation Algorithm for 2-Edge-Connected Spanning Subgraphs. Proc. 26th International Computing and Combinatorics Conference (COCOON’20) (2020).
38.
Steininger, M., Kobs, K., Zehe, A., Lautenschlager, F., Becker, M., Hotho, A.: MapLUR: Exploring a New Paradigm for Estimating Air Pollution Using Deep Learning on Map Images. ACM Trans. Spatial Algorithms Syst. 6, (2020).
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.
39.
Steininger, M., Abel, D., Ziegler, K., Krause, A., Paeth, H., Hotho, A.: Deep Learning for Climate Model Output Statistics. Tackling Climate Change with Machine Learning Workshop at NeurIPS 2020. (2020).
40.
Lautenschlager, F., Becker, M., Kobs, K., Steininger, M., Davidson, P., Krause, A., Hotho, A.: OpenLUR: Off-the-shelf air pollution modeling with open features and machine learning. Atmospheric Environment. 233, 117535 (2020).
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.
41.
Fischer, M., Kobs, K., Hotho, A.: NICER — Aesthetic Image Enhancement with Humans in the Loop. ACHI 2020: The Thirteenth International Conference on Advances in Computer-Human Interactions. S. 357–362 (2020).
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.
42.
Kobs, K., Potthast, M., Wiegmann, M., Zehe, A., Stein, B., Hotho, A.: Towards Predicting the Subscription Status of Twitch.tv Users. Proceedings of ECML-PKDD 2020 ChAT Discovery Challenge on Chat Analytics for Twitch. (2020).
43.
Schlör, D., Ring, M., Krause, A., Hotho, A.: Financial Fraud Detection with Improved Neural Arithmetic Logic Units. (2020).
44.
Zehe, A., Arns, J., Hettinger, L., Hotho, A.: HarryMotions – Classifying Relationships in Harry Potter based on Emotion Analysis. 5th SwissText & 16th KONVENS Joint Conference (2020).
45.
Schlör, D., Zehe, A., Kobs, K., Veseli, B., Westermeier, F., Brübach, L., Roth, D., Latoschik, M.E., Hotho, A.: Improving Sentiment Analysis with Biofeedback Data. Proceedings of LREC2020 Workshop ``People in language, vision and the mind’’ (ONION2020). S. 28–33. European Language Resources Association (ELRA), Marseille, France (2020).
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.
46.
Omeliyanenko, J., Zehe, A., Hettinger, L., Hotho, A.: LM4KG: Improving Common Sense Knowledge Graphs with Language Models. In: Pan, J.Z., Tamma, V., d’Amato, C., Janowicz, K., Fu, B., Polleres, A., Seneviratne, O., und Kagal, L. (hrsg.) The Semantic Web -- ISWC 2020. S. 456–473. Springer International Publishing, Cham (2020).
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?
47.
Wienrich, C., Döllinger, N.I., Hein, R.: Mind the Gap: A Framework (BehaveFIT) Guiding The Use of Immersive Technologies in Behavior Change Processes. arXiv preprint arXiv:2012.10912. (2020).
48.
Tönsing, L.: Prediction of YouTube’s Chunk Request from Monitored Encrypted Network Traffic, (2020).
49.
Seufert, M.: Statistical Methods and Models based on Quality of Experience Distributions. Quality and User Experience. 6, (2020).
Due to biased assumptions on the underlying ordinal rating scale in subjective Quality of Experience (QoE) studies, Mean Opinion Score (MOS)-based evaluations provide results, which are hard to interpret and can be misleading. This paper proposes to consider the full QoE distribution for evaluating, reporting, and modeling QoE results instead of relying on MOS-based metrics derived from results based on ordinal rating scales. The QoE distribution can be represented in a concise way by using the parameters of a multinomial distribution without losing any information about the underlying QoE ratings, and even keeps backward compatibility with previous, biased MOS-based results. Considering QoE results as a realization of a multinomial distribution allows to rely on a well-established theoretical background, which enables meaningful evaluations also for ordinal rating scales. Moreover, QoE models based on QoE distributions keep detailed information from the results of a QoE study of a technical system, and thus, give an unprecedented richness of insights into the end users’ experience with the technical system. In this work, existing and novel statistical methods for QoE distributions are summarized and exemplary evaluations are outlined. Furthermore, using the novel concept of quality steps, simulative and analytical QoE models based on QoE distributions are presented and showcased. The goal is to demonstrate the fundamental advantages of considering QoE distributions over MOS-based evaluations if the underlying rating data is ordinal in nature.
50.
Borchert, K., Seufert, A., Gamboa, E., Hirth, M., Hoßfeld, T.: In vitro vs in vivo: does the study’s interface design influence crowdsourced video QoE?. Quality and User Experience. (2020).
Evaluating the Quality of Experience (QoE) of video streaming and its influence factors has become paramount for streaming providers, as they want to maintain high satisfaction for their customers. In this context, crowdsourced user studies became a valuable tool to evaluate different factors which can affect the perceived user experience on a large scale. In general, most of these crowdsourcing studies either use, what we refer to, as an in vivo or an in vitro interface design. In vivo design means that the study participant has to rate the QoE of a video that is embedded in an application similar to a real streaming service, e.g., YouTube or Netflix. In vitro design refers to a setting, in which the video stream is separated from a specific service and thus, the video plays on a plain background. Although these interface designs vary widely, the results are often compared and generalized. In this work, we use a crowdsourcing study to investigate the influence of three interface design alternatives, an in vitro and two in vivo designs with different levels of interactiveness, on the perceived video QoE. Contrary to our expectations, the results indicate that there is no significant influence of the study’s interface design in general on the video experience. Furthermore, we found that the in vivo design does not reduce the test takers’ attentiveness. However, we observed that participants who interacted with the test interface reported a higher video QoE than other groups.
51.
Dulny, A., Steininger, M., Lautenschlager, F., Krause, A., Hotho, A.: Evaluating the multi-task learning approach for land use regression modelling of air pollution. International Conference on Frontiers of Artificial Intelligence and Machine Learning. IASED (2020).
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.
52.
Borchert, K.: Estimating Quality of Experience of Enterprise Applications - A Crowdsourcing-based Approach, (2020).
53.
Tritscher, J., Ring, M., Schlr, D., Hettinger, L., Hotho, A.: Evaluation of Post-hoc XAI Approaches Through Synthetic Tabular Data. In: Helic, D., Leitner, G., Stettinger, M., Felfernig, A., und Ra’s, Z.W. (hrsg.) Foundations of Intelligent Systems. S. 422–430. Springer International Publishing, Cham (2020).
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.
54.
Kobs, K., Steininger, M., Zehe, A., Lautenschlager, F., Hotho, A.: SimLoss: Class Similarities in Cross Entropy. In: Helic, D., Leitner, G., Stettinger, M., Felfernig, A., und Ra’s, Z.W. (hrsg.) Foundations of Intelligent Systems. S. 431–439. Springer International Publishing, Cham (2020).
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
55.
Truman, S., Mammen, textbfSebastian von: An Integrated Design of World-in-Miniature Navigation in Virtual Reality. FDG ’20: Proceedings of the 15th International Conference on the Foundations of Digital Games. S. 1–9. ACM, Bugibba, Malta (2020).
Navigation is considered one of the most fundamental challenges in Virtual Reality (VR) and has been extensively researched [11]. The world-in-miniature (WIM) navigation metaphor allows users to travel in large-scale virtual environments (VEs) regardless of available physical space while maintaining a high-level overview of the VE. It relies on a hand-held, scaled-down duplicate of the entire VE, where the user’s current position is displayed, and an interface provided to introduce his/her next movements [17]. There are several extensions to deal with challenges of this navigation technique, e.g. scaling and scrolling [23]. In this work, a WIM is presented that integrates state-of-the-art research insights and incorporates additional features that became apparent during the integration process. These features are needed to improve user interactions and to provide both look-ahead and post-travel feedback. For instance, a novel occlusion handling feature hides the WIM geometry in a rounded space reaching from the user’s hand to his/her forearm. This allows the user to interact with occluded areas of the WIM such as buildings. Further extensions include different visualizations for occlusion handling, an interactive preview screen, post-travel feedback, automatic WIM customization, a unified diegetic UI design concerning WIM and user representation, and an adaptation of widely established gestures to control scaling and scrolling of the WIM. Overall, the presented WIM design integrates and extends state-of-the-art interaction tasks and visualization concepts to overcome open conceptual gaps and to provide a comprehensive practical solution for traveling in VR.
56.
Gray, N.: Evaluation of the Impact of Network Topologies and Applications on Synchronization Strategies of Distributed SDN Controllers, (2020).
Software-defined Networking (SDN) separates the data from the control plane and outsources the latter to a logically centralized instance called the controller. This enables a common view of the network topology and provides a central entity for network configurations. The controller is often distributed to impose no single point of failure and to balance the load efficiently among the physical controller instances. To maintain a consistent view of the network, synchronization between the individual controller instances is required. Controller implementations can be categorized by their utilized synchronization strategy and largely fall in one of two categories, namely horizontal or hierarchical. Goal of this thesis is to evaluate the impact of the network topology and running controller applications on the performance of the controller when using different synchronization mechanisms. As access to large SDN test beds is limited the investigations will be done by simulations within the OMNeT++ framework.
57.
Zimmerer, C., Heinrich, R., Fischbach, M., Lugrin, J.-L., Latoschik, M.E.: Computing Object Selection Difficulty in VR Using Run-Time Contextual Analysis. 26th ACM Symposium on Virtual Reality Software and Technology. Association for Computing Machinery, Virtual Event, Canada (2020).
This paper introduces a method for computing the difficulty of selection tasks in virtual environments using pointing metaphors by operationalizing an established human motor behavior model. In contrast to previous work, the difficulty is calculated automatically at run-time for arbitrary environments. We present and provide the implementation of our method within Unity 3D. The difficulty is computed based on a contextual analysis of spatial boundary conditions, i.e., target object size and shape, distance to the user, and occlusion. We believe our method will enable developers to build adaptive systems that automatically equip the user with the most appropriate selection technique according to the context. Further, it provides a standard metric to better evaluate and compare different selection techniques.
58.
Kindermann, P., Mchedlidze, T., Meulemans, W., Rutter, I.: Graph Drawing Contest Report. In: Auber, D. und Valtr, P. (hrsg.) Proc. 28th International Symposium on Graph Drawing and Network Visualization (GD’20) (2020).
59.
Zimmerer, C., Wolf, E., Wolf, S., Fischbach, M., Lugrin, J.-L., Latoschik, M.E.: Finally on Par?! Multimodal and Unimodal Interaction for Open Creative Design Tasks in Virtual Reality. Proceedings of the 2020 International Conference on Multimodal Interaction. S. 222–231. Association for Computing Machinery, Virtual Event, Netherlands (2020).
Multimodal Interfaces (MMIs) have been considered to provide promising interaction paradigms for Virtual Reality (VR) for some time. However, they are still far less common than unimodal interfaces (UMIs). This paper presents a summative user study comparing an MMI to a typical UMI for a design task in VR. We developed an application targeting creative 3D object manipulations, i.e., creating 3D objects and modifying typical object properties such as color or size. The associated open user task is based on the Torrence Tests of Creative Thinking. We compared a synergistic multimodal interface using speech-accompanied pointing/grabbing gestures with a more typical unimodal interface using a hierarchical radial menu to trigger actions on selected objects. Independent judges rated the creativity of the resulting products using the Consensual Assessment Technique. Additionally, we measured the creativity-promoting factors flow, usability, and presence. Our results show that the MMI performs on par with the UMI in all measurements despite its limited flexibility and reliability. These promising results demonstrate the technological maturity of MMIs and their potential to extend traditional interaction techniques in VR efficiently.
60.
Obremski, D., Wienrich, C., Carolus, A.: What the user’s voice tells us about UX - Analysing parameters of the voice as indicators of the User Experience of the usage of intelligent voice assistants. Gehalten auf der (2020).
61.
Supervisor: Metzger, F.: Exploratory Comparison of Contemporary Cloud Gaming Services, (2020).
Recently, Cloud Gaming services have once again risen in popularity, with more and more commercial offerings becoming available. Your job in this thesis would be to investigate and compare these services on a technical level and based on their network characteristics.
62.
Gray, N.: Evaluation of Offloading Firewall Rules with P4, (2020).
Contact: nicholas.gray@informatik.uni-wuerzburg.de Firewalls are classical network middle boxes, which are physically implemented within the data path. Following this design pattern causes the network to become inflexible and harder to scale as additional hardware devices have to be deployed whenever the limits of the current configuration are reached, resulting in a high maintenance and requisition cost. Network Function Virtualization (NFV) is a novel paradigm aiming to mitigate these drawbacks by shifting the function of hardware middle boxes to software programs run on Commodity of the Shelf servers. While increasing the network flexibility, it also imposes additional delays on the data path as the complete networking and software stack has to be traversed for each packet. P4 allows the execution and packet matching process on inexpensive, programmable networking cards at line rate. Goal of the thesis is the implementation and evaluation of a distributed stateful firewall based on P4.
63.
Mertinat, N.: Impact of Content Selection on Crowdsourced QoE Studies of HTTP Adaptive Streaming on Mobile Devices, (2020).
64.
Brefeld, U., Fromont, Élisa, Hotho, A., Knobbe, A.J., Maathuis, M.H., Robardet, C. hrsg.: Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part I. Springer (2020).
65.
Aliev, E.: Evaluation of the Impact of Web Cross-Traffic on TCP and QUIC Video Streaming, (2020).
66.
Büttner, J., Merz, C., von Mammen, S.: Horde Battle III or How to Dismantle a Swarm. 2020 IEEE Conference on Games (CoG). S. 640–641. IEEE, Osaka, Japan (2020).
In this demo paper, we present the design of a virtual reality (VR) first-person shooter (FPS) in which the player fends off waves of hostile flying swarm robots that took over the Earth. The purpose of this serious game is to train the player in understanding networks by learning how to dismantle them. We explain the play and game mechanics and the level designs tailored to provide an engaging experience and to re-enforce the network perspective of the swarm dynamics.
67.
Chaplick, S., Fleszar, K., Lipp, F., Ravsky, A., Verbitsky, O., Wolff, A.: Drawing Graphs on Few Lines and Few Planes. Journal of Computational Geometry. 11, 433–475 (2020).
We investigate the problem of drawing graphs in 2D and 3D such that their edges (or only their vertices) can be covered by few lines or planes. We insist on straight-line edges and crossing-free drawings. This problem has many relations to other challenging graph-drawing problems such as small-area or small-volume drawings, layered or track drawings, and drawing graphs with low visual complexity. While some facts about our problem are implicit in previous work, this is the first treatment of the problem in its full generality. Our contribution is as follows. par We show lower and upper bounds for the numbers of lines and planes needed for covering drawings of graphs in certain graph classes. In some cases our bounds are asymptotically tight; in some cases we are able to determine exact values. par We relate our parameters to standard combinatorial characteristics of graphs (such as the chromatic number, treewidth, or arboricity) and to parameters that have been studied in graph drawing (such as the track number or the number of segments appearing in a drawing). par We pay special attention to planar graphs. For example, we show that there are qplanar graphs that can be drawn in 3-space on asymptotically fewer lines than in the plane.
68.
Wodarczyk, S., Mammen, textbfSebastian von: Emergent Multiplayer Games. IEEE CoG ’20: Proceedings of the 2nd International Conference on Games. S. 33–40. IEEE, Osaka, Japan (2020).
69.
Bauer, A., Züfle, M., Herbst, N., Zehe, A., Hotho, A., Kounev, S.: Time Series Forecasting for Self-Aware Systems. Proceedings of the IEEE. 1–26 (2020).
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.
70.
Brefeld, U., Fromont, Élisa, Hotho, A., Knobbe, A.J., Maathuis, M.H., Robardet, C. hrsg.: Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Würzburg, Germany, September 16-20, 2019, Proceedings, Part II. Springer (2020).
71.
Ziegler, P., Mammen, textbfSebastian von: Generating Real-Time Strategy Heightmaps using Cellular Automata. FDG ’20: Proceedings of the 15th International Conference on the Foundations of Digital Games. ACM, Bugibba, Malta (2020).
This paper presents a new approach of heightmap generation for Real-Time Strategy games (RTS) based on Cellular Automata (CA) in the context of various established techniques. The proposed approach uses different CA rulesets to generate and modify maps for the RTS game Supreme Commander. To evaluate the quality of the generated maps, a survey was conducted asking 30 participants about map quality compared to user-generated maps. The participants rated the maps more balanced and novel but less aesthetically pleasing. The paper concludes with according future work propositions to improve the presented approach.
72.
Supervisor: Metzger, F.: Designing a Realistic Virtual AQM Testbench, (2020).
Many new queue management techniques and supporting technologies are currently being developed, standardized and rolled out. CoDel, Cake, PIE, DualPI just to name a few. Your job is to develop a portable virtual environment that allows for realistic tests and comparisons of these methods with automated test suits.
73.
Supervisor: Grigorjew, A.: Configuring Stream Aggregates with Timed Gates for Predictable Latency in Dynamic Networks, (2020).
Time Synchronization and Timed Gates provide a popular mechanism for extremely low deterministic latency in Time-Sensitive Networking. Achieving the lowest latencies possible requires complex per-stream configuration and network-wide knowledge. This type of preparation is not possible in dynamic networks with varying traffic requirements and changing traffic patterns. This work is focused on the preparation of aggregate schedules that prepare time slots for a specific amount of traffic beforehand, which can be dynamically used by streams during the operation of the network. The challenges include the modeling of dynamic network requirements and, based on that, an optimal configuration of the number and length of transmission windows for each priority of this aggregate scheduling model.
74.
Borchert, K., Seufert, M., Hildebrand, K., Hoßfeld, T.: QoE Assessment of Enterprise Applications based on Self-motivated Ratings. 12th International Conference on Quality of Multimedia Experience (QoMEX). , Athlone, Ireland (2020).
In most companies, enterprise applications, such as office products or databases, are heavily used by employees during work hours. Impairments and performance issues not only slow down business processes, but might also increase the frustration of the workforce. While Quality of Experience (QoE) has been widely studied for personal multimedia applications, such as video streaming, its application to the business usage domain is still in its infancy. Due to several reasons, e.g., the high complexity of IT infrastructure, classical QoE studies can hardly be transferred to business applications. These studies are often independent from the context of usage and actively poll ratings from their participants. This work contrasts the commonly used "pull" method for collecting user ratings with a self-motivated "push" approach. This approach is inspired by complaint systems, in which users can directly report problems with a technical system as soon as they notice them. Therefore, performance assessments of a business application from employees of a cooperating company are collected with both rating systems during a time span of 1.5 years. Besides the analysis of the interaction of users with the "push" system, differences between the two methods are discussed. Further, QoE models for the monitored business application are derived based on the self-motivated "push" ratings.
75.
Kargl, J.: Analysis of QoE Aspects of 3D Point Cloud Reduction, (2020).
76.
Moldovan, C., Loh, F., Seufert, M., Hoßfeld, T.: Optimizing HAS for 360-Degree Videos. 5th IEEE/IFIP International Workshop on Analytics for Network and Service Management (AnNet). , Budapest, Hungary (2020).
77.
Berani, S.: Fingerprinting Websites in Encrypted Network Traffic, (2020).
Contact/Kontakt: nikolas.wehner@informatik.uni-wuerzburg.de Auch als Praktikum verfügbar. Due to the encryption of network traffic and the introduction of DNS over HTTPS, the task of identifying websites based on the corresponding network traffic got difficult. Nonetheless, ISPs want to know which kind of traffic passes their networks in order to optimize the QoE of their customers. The goal of this work is to create a model which is able to reliably predict the called websites during a session. Therefore, reasonable features have to be extracted from the data and afterward they have to be ingested into selected Machine Learning algorithms.
78.
Chaplick, S., Kindermann, P., Löffler, A., Thiele, F., Wolff, A., Zaft, A., Zink, J.: Stick Graphs with and without Length Constraints. Journal of Graph Algorithms & Applications. 25 pages (2020).
Stick graphs are intersection graphs of horizontal and vertical line seg- ments that all touch a line of slope ���1 and lie above this line. De Luca et al. [GD'18] considered the recognition problem of stick graphs when no order is given (STICK), when the order of either one of the two sets is given (STICK\($_A$\)), and when the order of both sets is given (STICK\($_AB$\)). They showed how to solve STICK\($_AB$\) efficiently. par In this paper, we improve the running time of their algorithm, and we solve STICK\($_A$\) efficiently. Further, we consider variants of these problems where the lengths of the sticks are given as input. We show that these variants of STICK, STICK\($_A$\), and STICK\($_AB$\) are all NP-complete. On the positive side, we give an efficient solution for STICK\($_AB$\) with fixed stick lengths if there are no isolated vertices.
79.
Chaplick, S., Förster, H., Kryven, M., Wolff, A.: Drawing Graphs with Circular Arcs and Right-Angle Crossings. In: Albers, S. (hrsg.) Proc. 17th Scand. Symp. and Workshops on Algorithm Theory (SWAT’20). S. 21:1–. Schloss Dagstuhl~-- Leibniz-Zentrum für Informatik (2020).
80.
Supervisor: Loh, F.: Demand Study and Performance Evaluation of Future Smart City Networks, (2020).
Zukünftige Netze müssen durch den steigenden Einsatz von Smart City Technologien sowie einer höheren Anzahl an Geräten im Netz durch vermehrten Sensor- und IoT Einsatz angepasst werden. Dafür soll in der Arbeit untersucht werden, wie sich in Zukunft die Verkehrslast verändert. Dies kann mittels Messung, Simulation, oder Analyse betrachtet werden.
81.
Chaplick, S., van Dijk, T.C., Kryven, M., Park, J.-W., Ravsky, A., Wolff, A.: Bundled Crossings Revisited. Journal of Graph Algorithms & Applications. 35 pages (2020).
An effective way to reduce clutter in a graph drawing that has (many) crossings is to group edges that travel in parallel into bundles. Each edge can participate in many such bundles. Any crossing in this bundled graph occurs between two bundles, i.e., as a bundled crossing. We consider the problem of bundled crossing minimization: A graph is given and the goal is to find a bundled drawing with at most \($k$\) bundled crossings. We show that the problem is NP-hard when we require a simple drawing. Our main result is an FPT algorithm (in \($k$\)) for simple circular layouts where vertices must be placed on a circle and edges must be drawn inside the circle. These results make use of the connection between bundled crossings and graph genus. We also consider bundling crossings in a given drawing, in particular for storyline visualizations.
82.
Hildebrand, K.: Impact of Rendering Times on the Perceived Loading Speed of Web Pages, (2020).
Contact/Kontakt: michael.seufert@uni-wuerzburg.de Title: Impact of Rendering Times on the Perceived Loading Speed of Web Pages Abstract: The most dominant QoE (Quality of Experience, i.e., subjectively perceived quality) influence factor of web browsing is the page load time. However, pages do not load in a single event, but different parts of the web page are displayed at different rendering times. In this work, a study shall be conducted to investigate the impact of rendering times on the perceived loading speed of web pages. The results of the study shall be evaluated to derive recommendations for networking as well as rendering aspects of web browsing. Titel: Einfluss von Anzeigezeiten auf die wahrgenomme Ladegeschwindigkeit von Webseiten Inhalt: Der wichtigste Einflussfaktor auf die QoE (Quality of Experience, d.h. subjektiv empfundene Dienstgüte) beim Webbrowsing ist die Seitenladezeit. Jedoch laden Webseiten nicht auf einmal, sondern verschiedene Teile der Seite werden zu unterschiedlichen Zeitpunkten angezeigt. In dieser Arbeit soll eine Studie durchgeführt werden, mit der der Einfluss dieser Anzeigezeiten auf die wahrgenommene Ladegeschwindigkeit untersucht werden kann. Die Ergebnisse der Studie sollen ausgewehrtet werden, um Empfehlungen für Netzwerk- und Renderingaspekte beim Webbrowsen abzuleiten.
83.
Loh, F., Wamser, F., Moldovan, C., Zeidler, B., Tsilimantos, D., Valentin, S., Hoßfeld, T.: Is the Uplink Enough? Estimating Video Stalls from Encrypted Network Traffic. 32th IEEE/IFIP Network Operations and Management Symposium. , Budapest, Hungary (2020).
Today’s traffic projections speak of almost 58 % video traffic across the Internet. Nearly all video traffic is encrypted, accounting for more than 50 % encrypted traffic worldwide. To analyze video traffic today, or even estimate its quality in the network, a deep look into the traffic characteristics has to be done. But then, important quality metrics from the traffic behavior can be derived. Based on extensive measurements we show in this work how to measure and estimate video stalls for mobile adaptive streaming. The underlying dataset includes more than 900 hours of video footage from the native YouTube app, measured over 18 different videos in 56 network scenarios in two cities in Europe. We outline a possible approach to estimate the video playback buffer size based on uplink video chunk requests in real-time to break down the video stalls. This work is intended as a tool for network operators to receive further knowledge of the characteristics of video streaming traffic to quantify the most important QoE degradation factors of one of the most important applications today.
84.
Korger, A., Baumeister, J.: Case-Based Generation of Regulatory Documents and their Semantic Relatedness. Proceedings of the Future of Information and Communications Conference (FICC) 2020. Springer, San Francisco (2020).
85.
Walter, J., Zink, J., Baumeister, J., Wolff, A.: Layered Drawing of Undirected Graphs with Generalized Port Constraints. In: Auber, D. und Valtr, P. (hrsg.) Proc. 28th Int. Symp. Graph Drawing & Network Vis. (GD’20). Springer-Verlag (2020).
86.
Halloway, M.: Investigating the Influence of App Browsing Delays on the QoE of Music Streaming, (2020).
87.
Fehler, H.: Comparison of Web QoE Algorithms on Different Devices, (2020).
Contact/Kontakt: nikolas.wehner@informatik.uni-wuerzburg.de Impairments in the Web QoE, e.g., low page load times, can lead to annoyed users which in the end can result in revenue losses for companies. In literature, several algorithm exist for predicting the Web QoE, e.g. Speed Index, Above the Fold, etc. Based on data from a measurement study, these algorithms shall be evaluated and compared in this work. In particular, a comparison of the metrics among the devices (Smartphone, Tablet, Desktop PC), which were used to generate the measurement data, is of interest.
88.
Supervisor: Loh, F.: A Simulative Approach to Study the Coverage of LoRaWAN in Würzburg, (2020).
In der Masterarbeit soll anhand von bereits vorliegenden Daten die Abdeckung eines LoRaWAN Netzes in Würzburg simuliert werden. Dabei sollen verschiedene Sensor- sowie Verkehrsmuster simuliert werden und der Einfluss auf eine Smart City untersucht werden.
89.
Wehner, N.: Machine Learning-based Real-time Estimation of Quality of Experience from Encrypted Video Streaming Traffic. , 1st KuVS Workshop on Machine Learning for Networking, Munich, Germany (2020).
90.
Peng, D., Wolff, A., Haunert, J.-H.: Finding Optimal Sequences for Area Aggregation---A\($^\star$\) vs. Integer Linear Programming. ACM Transactions on Spatial Algorithms and Systems. 7, (2020).
To provide users with maps of different scales and to allow them to zoom in and out without losing context, automatic methods for map generalization are needed. We approach this problem for land-cover maps. Given two land-cover maps at two different scales, we want to find a sequence of small incremental changes that gradually transforms one map into the other. We assume that the two input maps consist of polygons, each of which belongs to a given land-cover type. Every polygon on the smaller-scale map is the union of a set of adjacent polygons on the larger-scale map. par In each step of the computed sequence, the smallest area is merged with one of its neighbors. We do not select that neighbor according to a prescribed rule but compute the whole sequence of pairwise merges at once, based on global optimization. We have proved that this problem is NP-hard. We formalize this optimization problem as that of finding a shortest path in a (very large) graph. We present the A\($^\star$\) algorithm and integer linear programming to solve this optimization problem. To avoid long computing times, we allow the two methods to return non-optimal results. In addition, we present a greedy algorithm as a benchmark. We tested the three methods with a dataset of the official German topographic database ATKIS. Our main result is that A\($^\star$\) finds optimal aggregation sequences for more instances than the other two methods within a given time frame.
91.
Raffeck, S.: Towards Modeling Energy Performance of LoRaWAN Transmissions: A Verification of Theoretical Predictions, (2020).
92.
Supervisor: Loh, F.: Studying the Scalability of LoRaWAN with Omnet++, (2020).
93.
Wamser, F., Alay, Özgü, Metzger, F., Valentin, S.: IJNM Special Issue - Editorial - QoE-centric Analysis and Management of Communication Networks. International Journal of Network Management. Special Issue: QoE-centric Analysis and Management of Communication Networks, (2020).
The heterogeneity and variability of Internet applications has increased considerably in recent years. Applications such as video streaming are responsible for a large part of data traffic on the Internet. Internet telephony and video conferencing systems have become part of our daily lives. At the same time, the Internet of Things is striving to exceed previous expectations regarding the number of devices. Furthermore, the proliferation of video games, virtual reality applications, and 360° video applications is increasing. All this leads to specific but different requirements from applications to frameworks, service platforms, and networks. For each service, users desire special service criteria, such as smooth interactivity, fast downloads, high availability, or extensive content. Such requirements can usually be summarized under the term Quality of Experience, i.e., the overall satisfaction of a user with the system cur- rently in use. In the age of big data and dynamic networks, Quality of Experience is still looking for its place, and good solutions are in high demand. This Special Issue addresses the latest advances and challenges in analysis, design, modeling, measurement, and performance evaluation of Quality of Experience and Quality of Experience-oriented metrics and management.
94.
Hirth, M., Borchert, K., de Moor, K., Borst, V., Hoßfeld, T.: Personal Task Design Preferences of Crowdworkers. 12th International Conference on Quality of Multimedia Experience (QoMEX). , Athlone, Ireland (2020).
95.
Oyanagi, A., Narumi, T., Lugrin, J.-L., Ando, H., Ohmura, R.: Reducing the Fear of Height by Inducing Proteus Effect of a Dragon Avatar. Journal of the Virtual Reality Society of Japan (2020).
96.
Gall, D., Latoschik, M.E.: Visual angle modulates affective responses to audiovisual stimuli. Computers in Human Behavior. 109, 106346 (2020).
What we see influences our emotions. Technology often mediates the visual content we perceive. Visual angle is an essential parameter of how we see such content. It operationalizes visible properties of human-computer interfaces. However, we know little about the content-independent effect of visual angle on emotional responses to audiovisual stimuli. We show that visual angle alone affects emotional responses to audiovisual features, independent of object perception. We conducted a 2 x 2 x 3 factorial repeated-measures experiment with 143 undergraduate students. We simultaneously presented monochrome rectangles with pure tones and assessed valence, arousal, and dominance. In the high visual angle condition, arousal increased, valence and dominance decreased, and lightness modulated arousal. In the low visual angle condition, pitch modulated arousal, and lightness affected valence. Visual angle weights the affective relevance of perception modalities independent of spatial representations. Visual angle serves as an early-stage perceptual feature for organizing emotional responses. Control of this presentation layer allows for provoking or avoiding emotional response where intended.
97.
Hoßfeld, T., Heegaard, P.E., Varela, M., Skorin-Kapov, L., Fiedler, M.: From QoS Distributions to QoE Distributions: a System’s Perspective. 4th International Workshop on Quality of Experience Management (QoE Management 2020), featured by IEEE Conference on Network Softwarization (IEEE NetSoft 2020). , Ghent, Belgium (2020).
In the context of QoE management, network and service providers commonly rely on models that map system QoS conditions (e.g., system response time, packet loss, etc.) to estimated end user QoE values. Observable QoS conditions in the system may be assumed to follow a certain distribution, meaning that different end users will experience different conditions. On the other hand, drawing from the results of subjective user studies, we know that user diversity leads to distributions of user scores for any given test conditions (in this case referring to the QoS parameters of interest). Our previous studies have shown that to correctly derive various QoE metrics (e.g., Mean Opinion Score (MOS), quantiles, probability of users rating 'good or better', etc.) in a system under given conditions, there is a need to consider rating distributions obtained from user studies, which are often times not available. In the paper we extend these findings to show how to approximate user rating distributions given a QoS-to-MOS mapping function and second order statistics. Such a user rating distribution may then be combined with a QoS distribution observed in a system to finally derive corresponding distributions of QoE scores. Scripts are provided for the approximating a QoE distribution with a Beta distribution for given MOS and SOS parameter at https://github.com/hossfeld/approx-qoe-distribution.
98.
Wißmann, N., Mišiak, M., Fuhrmann, A., Latoschik, M.E.: Accelerated Stereo Rendering with Hybrid Reprojection-Based Rasterization and Adaptive Ray-Tracing. Proceedings of the 27th IEEE Virtual Reality conference (IEEE VR ’20) (2020).
Stereoscopic rendering is a prominent feature of virtual reality ap- plications to generate depth cues and to provide depth perception in the virtual world. However, straight-forward stereo rendering methods usually are expensive since they render the scene from two eye-points which in general doubles the frame times. This is particularly problematic since virtual reality sets high requirements for real-time capabilities and image resolution. Hence, this paper presents a hybrid rendering system that combines classic rasteriza- tion and real-time ray-tracing to accelerate stereoscopic rendering. The system reprojects the pre-rendered left half of the stereo image pair into the right perspective using a forward grid warping technique and identifies resulting reprojection errors, which are then efficiently resolved by adaptive real-time ray-tracing. A final analysis shows that the system achieves a significant performance gain, has a neg- ligible quality impact, and is suitable even for higher rendering resolutions.
99.
Hoßfeld, T., Wunderer, S., Beyer, A., Hall, A., Schwind, A., Gassner, C., Guillemin, F., Wamser, F., Wascinski, K., Hirth, M., Seufert, M., Casas, P., Tran-Gia, P., Robitza, W., Wascinski, W., Ben Houidi, Z.: White Paper on Crowdsourced Network and QoE Measurements – Definitions, Use Cases and Challenges. Würzburg (2020).
The goal of the white paper at hand is as follows. The definitions of the terms build a framework for discussions around the hype topic ‘crowdsourcing’. This serves as a basis for differentiation and a consistent view from different perspectives on crowdsourced network measurements, with the goal to provide a commonly accepted definition in the community. The focus is on the context of mobile and fixed network operators, but also on measurements of different layers (network, application, user layer). In addition, the white paper shows the value of crowdsourcing for selected use cases, e.g., to improve QoE or regulatory issues. Finally, the major challenges and issues for researchers and practitioners are highlighted. This white paper is the outcome of the Würzburg seminar on “Crowdsourced Network and QoE Measurements” which took place from 25-26 September 2019 in Würzburg, Germany. International experts were invited from industry and academia. They are well known in their communities, having different backgrounds in crowdsourcing, mobile networks, network measurements, network performance, Quality of Service (QoS), and Quality of Experience (QoE). The discussions in the seminar focused on how crowdsourcing will support vendors, operators, and regulators to determine the Quality of Experience in new 5G networks that enable various new applications and network architectures. As a result of the discussions, the need for a white paper manifested, with the goal of providing a scientific discussion of the terms “crowdsourced network measurements” and “crowdsourced QoE measurements”, describing relevant use cases for such crowdsourced data, and its underlying challenges. During the seminar, those main topics were identified, intensively discussed in break-out groups, and brought back into the plenum several times. The outcome of the seminar is this white paper at hand which is – to our knowledge – the first one covering the topic of crowdsourced network and QoE measurements.
100.
Plesker, F.: Simulation of a 5G Network Infrastructure, (2020).
101.
Wehner, N., Mertinat, N., Seufert, M., Hoßfeld, T.: Studying the Impact of the Content Selection Method on the Video QoE on Mobile Devices. 12th International Conference on Quality of Multimedia Experience (QoMEX). , Athlone, Ireland (2020).
When conducting video QoE studies, participants are usually asked to rate the QoE of prepared test videos. However, participants are given no choice to select content, which they like or in which they are interested. This may cause annoyance or frustration when conducting the QoE study, which eventually might affect the QoE results of the study. The consequent question is whether the content liking has a direct impact on the submitted ratings by the participants and whether the freedom of choosing the video content in QoE studies results in better ratings. To investigate this research question, MCA, an existing framework for crowdsourced video testing, is extended and used in a pilot field study. MCA runs on mobile devices and allows users to watch the tested conditions of a QoE study within video content of their interest. The results of a QoE study with individual and dynamic content selection are compared to a QoE study with pre-selected contents. Moreover, this work includes a comparison to a previous QoE study for validation. As the previous study was conducted on desktop PCs, the MCA study further allows to identify differences in the stalling perception between studies on desktop PCs and mobile devices.
102.
Schlör, D., Zehe, A., Kobs, K., Veseli, B., Westermeier, F., Brübach, L., Roth, D., Latoschik, M.E., Hotho, A.: Improving Sentiment Analysis with Biofeedback Data. Proceedings of the Workshop on peOple in laNguage, vIsiOn and the miNd (ONION) (2020).
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.
103.
Mal, D.: [DC] The Impact of Social Interactions on an Embodied Individual’s Self-perception in Virtual Environments. 2020 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW). S. 545–546. IEEE, Atlanta, GA, USA, USA (2020).
In shared immersive virtual reality, users can interact with other participants and experience them as being present in the environment. Thereby different aspects of the respective interaction partners can have an impact on the perceived quality of the communication and possibly also the self-perception of an embodied user. This paper describes various factors the author aims to investigate during his doctoral studies. As the research area of embodied social interactions is broad, relevant factors and concrete research questions have been identified to investigate how social contact with one or multiple other persons may affect one’s self-perception, behavior and her or his relationship with the others while being embodied in a virtual environment.
104.
Ripka, G., Tiede, J., Grafe, S., Latoschik, M.E.: Teaching and Learning Processes in Immersive VR – Comparing Expectations of Preservice Teachers and Teacher Educators. Society for Information Technology & Teacher Education (SITE) International Conference (2020).
The usage of VR in higher education is not uncommon anymore. However, concepts are mainly still focusing on technical rather than pedagogical aspects of VR in the classroom. The exploration of the expectations of teacher educators as well as of preservice teachers appears indispensable (1) to achieve a sound understanding of requirements, (2) to identify potential design spaces, and finally (3) to create and to derive suitable pedagogical approaches for VR in initial teacher education. This paper presents results of guideline-based qualitative interviews comparing the expectations of teacher educators and of preservice teachers regarding teaching and learning in immersive virtual learning environments. The results showed that preservice teachers and teacher educators expect VR to enrich classes through interactive engagement in situations that would otherwise be too costly or dangerous. Regarding the design, teacher educators put the emphasis on functionality. Student teachers emphasized that they do not want to miss social interactions with their peers. Furthermore, both groups stated preferred modes of collaboration and interaction taking into account the characteristics of a virtual learning surrounding such as being able to use diverse learning spaces for group work. Interviewees agreed on two vital factors for effective learning and teaching processes: flexibility and the possibility of customization considering technical properties that are to deal with. Apart from this, preservice teachers emphasized strongly their worries about data usage and the ethics regarding using avatars and agents for representation.
105.
Evans, W., Gethner, E., Spalding-Jamieson, J., Wolff, A.: Angle Covers: Algorithms and Complexity. In: Rahman, S., Sadakane, K., und Sung, W.-K. (hrsg.) Proc. 14th Int. Workshop Algorithms Comput. (WALCOM’20). S. 94–106. Springer-Verlag (2020).
106.
Spoerhase, J., Storandt, S., Zink, J.: Simplification of Polyline Bundles. Proc. 17th Scandinavian Symposium and Workshops on Algorithm Theory (SWAT’20) (2020).
107.
Schwind, A., Moldovan, C., Janiak, T., Dworschak, N.-D., Hoßfeld, T.: Don’t Stop the Music: Crowdsourced QoE Assessment of Music Streaming with Stalling. 12th International Conference on Quality of Multimedia Experience (QoMEX). , Athlone, Ireland (2020).
Streaming made a lasting effect on the way our society consumes media in the last decade. While due to streaming the way we listen to music and podcasts has changed drastically, there are very few studies about its Quality of Experience (QoE) and possible influence factors. From video QoE studies, we know that, for example, undesirable stops of the stream (stalling events) have a significant impact on QoE. However, the way in which music and video streaming is consumed differs significantly, as music is often played in the background, and thus, the influence of stalling could be significantly different. Thus, this work evaluates the impact of stalling on music streaming QoE. Therefore, we conduct two crowdsourced user studies: In the first study, users have to rate four songs with different stalling patterns and evaluate the degree of impairments. Afterwards, we compare the ratings to the results of a lab study and show that they are highly correlated, and that crowdsourcing is a suitable way of measuring music streaming QoE. In addition, we conduct a second crowdsourcing study to investigate the influence of the user's attentiveness on QoE. Here, participants have to listen to one song with two stalling events, while one half of them had to transcribe a handwritten text with music playing in the background. The attentiveness shows no influence on the perceived streaming quality, but it shows a significant influence on the perceived quality degradation due to stalling events. Furthermore, considerably more stalling events were missed for workers who focused on the transcription. These results are an important step towards establishing new methods for investigating QoE in multimedia.
108.
Supervisor: Loh, F.: Analyzing the Streaming Behavior of the Amazon Echo Show, (2020).
In der Arbeit soll das Streamingverhalten des Amazon Echo Shows gemessen und analysiert werden. Hauptsächlich soll der Einfluss von Bandbreitenschwankungen genauer betrachtet werden.
109.
Supervisor: Loh, F.: Use Case Study of the Transmission Behavior of an Application Using LoRaWAN, (2020).
Im Praktikum soll LoRaWAN näher betrachtet werden. Hier stellt sich die Frage, wie sich z.B. eine höhere Sendelast bzw. höhere Anforderungen bestimmter Anwendungen auf die Übertragung auswirken. Das kann mittels einer Simulation oder einer Analyse im Praktikum näher betrachtet werden.
110.
Fischer, E., Zoller, D., Dallmann, A., Hotho, A.: Integrating Keywords into BERT4Rec for Sequential Recommendation. KI 2020: Advances in Artificial Intelligence. S. 275–282. Springer International Publishing, Cham (2020).
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.
111.
Arslanian, A.: QoE Study for Compatible Video and Web Browsing QoE Models, (2020).
Contact: nikolas.wehner@uni-wuerzburg.de Abstract: Video streaming and web browsing are the dominant applications in the Internet today. Quality of Experience (QoE) is a concept introduced to describe the delight or frustration with a networked service. For both video streaming and web browsing, several QoE models have been proposed in literature. However, one persisting problem is the incompatiblity of these QoE models. As a consequence, modelling QoE for sessions including both video and web is an impossible task. Goal of this work is to design, implement, and conduct a QoE study which investigates both video streaming and web browsing within a session. The obtained results of the QoE study are then evaluated in order to propose compatible QoE models.
112.
Korger, A., Baumeister, J.: Construction of a Corpus for the Evaluation of Textual Case-based Reasoning Architectures. LWDA 2020 -- Lernen, Wissen, Daten, Analysen (2020).
113.
Oberdörfer, S., Elsässer, A., Schraudt, D., Grafe, S., Latoschik, M.E.: Horst – The Teaching Frog: Learning the Anatomy of a Frog Using Tangible AR. Proceedings of the 2020 Mensch und Computer Conference (MuC ’20). 303–307 (2020).
114.
Schlör, D., Ring, M., Hotho, A.: iNALU: Improved Neural Arithmetic Logic Unit. Frontiers in Artificial Intelligence. 3, 71 (2020).
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.
115.
Ziebell, P., Stümpfig, J., Eidel, M., Kleih, S.C., Kübler, A., Latoschik, M.E., Halder, S.: Stimulus modality influences session-to-session transfer of training effects in auditory and tactile streaming-based P300 brain–computer interfaces. Scientific Reports. 10, 11873-- (2020).
Despite recent successes, patients suffering from locked-in syndrome (LIS) still struggle to communicate using vision-independent brain–computer interfaces (BCIs). In this study, we compared auditory and tactile BCIs, regarding training effects and cross-stimulus-modality transfer effects, when switching between stimulus modalities. We utilized a streaming-based P300 BCI, which was developed as a low workload approach to prevent potential BCI-inefficiency. We randomly assigned 20 healthy participants to two groups. The participants received three sessions of training either using an auditory BCI or using a tactile BCI. In an additional fourth session, BCI versions were switched to explore possible cross-stimulus-modality transfer effects. Both BCI versions could be operated successfully in the first session by the majority of the participants, with the tactile BCI being experienced as more intuitive. Significant training effects were found mostly in the auditory BCI group and strong evidence for a cross-stimulus-modality transfer occurred for the auditory training group that switched to the tactile version but not vice versa. All participants were able to control at least one BCI version, suggesting that the investigated paradigms are generally feasible and merit further research into their applicability with LIS end-users. Individual preferences regarding stimulus modality should be considered.
116.
Wehner, N., Seufert, M., Egger-Lampl, S., Gardlo, B., Casas, P., Schatz, R.: Scoring High: Analysis and Prediction of Viewer Behavior and Engagement in the Context of 2018 FIFA WC Live Streaming. Proceedings of the 28th ACM International Conference on Multimedia (MM). , Seattle, WA, USA (2020).
117.
Niebling, F., Bruschke, J., Messemer, H., Wacker, M., von Mammen, S.: Analyzing Spatial Distribution of Photographs in Cultural Heritage Applications. In: Liarokapis, F., Voulodimos, A., Doulamis, N., und Doulamis, A. (hrsg.) Visual Computing for Cultural Heritage. S. 391–408. Springer, Cham (2020).
118.
Wehner, N., Mertinat, N., Seufert, M., Hoßfeld, T.: Studying the Impact of the Content Selection Method on the Video QoE on Mobile Devices, (2020).
When conducting video QoE studies, participants are usually asked to rate the QoE of prepared test videos. However, participants are given no choice to select content, which they like or in which they are interested. This may cause annoyance or frustration when conducting the QoE study, which eventually might affect the QoE results of the study. The consequent question is whether the content liking has a direct impact on the submitted ratings by the participants and whether the freedom of choosing the video content in QoE studies results in better ratings. To investigate this research question, MCA, an existing framework for crowdsourced video testing, is extended and used in a pilot field study. MCA runs on mobile devices and allows users to watch the tested conditions of a QoE study within video content of their interest. The results of a QoE study with individual and dynamic content selection are compared to a QoE study with pre-selected contents. Moreover, this work includes a comparison to a previous QoE study for validation. As the previous study was conducted on desktop PCs, the MCA study further allows to identify differences in the stalling perception between studies on desktop PCs and mobile devices.
119.
Donnermann, M., Schaper, P., Lugrin, B.: Integrating a Social Robot in Higher Education – A Field Study. 29th IEEE International Symposium on Robot and Human Interactive Communication. S. 573–579 (2020).
120.
Supervisor: Metzger, F.: Design and Evaluation of a Cloud Gaming Crowdsourcing Study, (2020).
The second wave of cloud gaming services is currently gaining traction. But we are still not sure of the subjective experience of games streamed over these services when compared to them running locally. Your task in this work would be to design the methodology for a crowdsourcing QoE assessment for cloud gaming, execute it and subsequently evaluate its results.
121.
Wolf, E., Döllinger, N., Mal, D., Wienrich, C., Botsch, M., Latoschik, M.E.: Body Weight Perception of Females using Photorealistic Avatars in Virtual and Augmented Reality. 2020 IEEE International Symposium on Mixed and Augmented Reality (ISMAR) (2020).
The appearance of avatars can potentially alter changes in their users' perception and behavior. Based on this finding, approaches to support the therapy of body perception disturbances in eating or body weight disorders by mixed reality (MR) systems gain in importance. However, the methodological heterogeneity of previous research has made it difficult to assess the suitability of different MR systems for therapeutic use in these areas. The effects of MR system properties and related psychometric factors on body-related perceptions have so far remained unclear. We developed an interactive virtual mirror embodiment application to investigate the differences between an augmented reality see-through head-mounted-display (HMD) and a virtual reality HMD on the before-mentioned factors. Additionally, we considered the influence of the participant's body-mass-index (BMI) and the BMI difference between participants and their avatars on the estimations. The 54 normal-weight female participants significantly underestimated the weight of their photorealistic, generic avatar in both conditions. Body weight estimations were significantly predicted by the participants' BMI and the BMI difference. We also observed partially significant differences in presence and tendencies for differences in virtual body ownership between the systems. Our results offer new insights into the relationships of body weight perception in different MR environments and provide new perspectives for the development of therapeutic applications.
122.
Supervisor: Metzger, F.: Vehicular and Wireless Traffic Simulation and Optimization under Urban Mobility, (2020).
Modern cities have complex interactions of vehicular, pedestrian and wireless networks. Any communal wireless network should strive to fullfil these demands. Your task would be to create a model of the city of Würzburg in SUMO (Simulation of Urban MObility) and OMNeT++ in order to evaluate Würzburg's (wireless) traffic demands and subsequently optimize them.
123.
Seufert, M.: Different Points of View: Impact of 3D Point Cloud Reduction on QoE of Rendered Images. , 12th International Conference on Quality of Multimedia Experience (QoMEX), Athlone, Ireland (virtual) (2020).
124.
Wassermann, S., Casas, P., Seufert, M., Wehner, N., Schüler, J., Hoßfeld, T.: How Good is your Mobile (Web) Surfing? SpeedIndex Inference from Encrypted Traffic, (2020).
We address the problem ofWeb QoE monitoring, in particular Speed Index (SI), from the Internet Service Provider (ISP) perspective, relying on in-network, passive measurements. Given the wide adoption of end-to-end encryption, we resort to machine-learning models to infer the SI of individual web-page loading sessions, using as input only packet-level data. Our study targets the analysis of SI in mobile devices, including smartphones and tablets. To the best of our knowledge, this is the first paper addressing the inference of SI from encrypted network traffic in mobile devices.
125.
Davidson, P., Düking, P., Zinner, C., Sperlich, B., Hotho, A.: Smartwatch-Derived Data and Machine Learning Algorithms Estimate Classes of Ratings of Perceived Exertion in Runners: A Pilot Study. Sensors. (2020).
126.
Landeck, M., Unruh, F., Lugrin, J.-L., Latoschik, M.E.: Metachron: A framework for time perception research in VR. Proceedings of the 26th ACM Conference on Virtual Reality Software and Technology (2020).
127.
Ganal, E., Bartl, A., Westermeier, F., Roth, D., Latoschik, M.E.: Developing a Study Design on the Effects of Different Motion Tracking Approaches on the User Embodiment in Virtual Reality. In: Hansen, C., Nürnberger, A., und Preim, B. (hrsg.) Mensch und Computer 2020. Gesellschaft für Informatik e.V (2020).
128.
Arseneva, E., Kleist, L., Klemz, B., Löffler, M., Schulz, A., Vogtenhuber, B., Wolff, A.: Representing Graphs by Polygons with Edge Contacts in 3D. In: Chaplick, S., Kindermann, P., und Wolff, A. (hrsg.) Proc. 36th European Workshop on Computational Geometry (EuroCG’20). S. 53:1– (2020).
129.
Hessenauer, L.: Classification of IoT Signaling Incidents, (2020).
130.
Kaufmann, M., Kratochvil, J., Lipp, F., Montecchiani, F., Raftopoulou, C., Valtr, P.: The Stub Resolution of 1-Planar Graphs. In: Rahman, M.S., Sadakane, K., und Sung, W.-K. (hrsg.) WALCOM: Algorithms and Computation. S. 170–182. Springer (2020).
The resolution of a drawing plays a crucial role when defining criteria for its quality. In the past, grid resolution, edge-length resolution, angular resolution and crossing resolution have been investigated. In this paper, we investigate the stub resolution, a recently introduced criterion for nonplanar drawings. A crossed edge is divided into parts, called stubs, which should not be too short for the sake of readability. Thus, the stub resolution of a drawing is defined as the minimum ratio between the length of a stub and the length of the entire edge, over all the edges of the drawing. We consider 1-planar graphs and we explore scenarios in which near optimal stub resolution, i.e. arbitrarily close to \\($\$\)\backslashfrac\1\\2\\\($\$\), can be obtained in drawings with zero, one, or two bends per edge, as well as further resolution criteria, such as angular and crossing resolution. In particular, our main contributions are as follows: (i) Every 1-planar graph with independent crossing edges has a straight-line drawing with near optimal stub resolution; (ii) Every 1-planar graph has a 1-bend drawing with near optimal stub resolution.
131.
Davidson, P., Steininger, M., Lautenschlager, F., Kobs, K., Krause, A., Hotho, A.: Anomaly Detection in Beehives using Deep Recurrent Autoencoders. Proceedings of the 9th International Conference on Sensor Networks (SENSORNETS 2020). S. 142–149. SCITEPRESS – Science and Technology Publications, Lda (2020).
132.
Byrka, J., Lewandowski, M., Meesum, S.M., Spoerhase, J., Uniyal, S.: PTAS for Steiner Tree on Map Graphs. Proc. 14th Latin American Theoretical Informatics Symposium (LATIN’20) (2020).
133.
Oberdörfer, S., Heidrich, D., Latoschik, M.E.: Think Twice: The Influence of Immersion on Decision Making during Gambling in Virtual Reality. Proceedings of the 27th IEEE Virtual Reality conference (VR ’20). S. 483–492. IEEE, Atlanta, USA (2020).
Immersive Virtual Reality (VR) is increasingly being explored as an alternative medium for gambling games to attract players. Typically, gambling games try to impair a player’s decision making, usually for the disadvantage of the players’ financial outcome. An impaired decision making results in the inability to differentiate between advantageous and disadvantageous options. We investigated if and how immersion impacts decision making using a VR-based realization of the Iowa Gambling Task (IGT) to pinpoint potential risks and effects of gambling in VR. During the IGT, subjects are challenged to draw cards from four different decks of which two are advantageous. The selections made serve as a measure of a participant’s decision making during the task. In a novel user study, we compared the effects of immersion on decision making between a low-immersive desktop-3D-based IGT realization and a high immersive VR version. Our results revealed significantly more disadvantageous decisions when playing the immersive VR version. This indicates an impair- ing effect of immersion on simulated real life decision making and provides empirical evidence for a high risk potential of gambling games targeting immersive VR.
134.
Bucher, K., Oberdörfer, S., Grafe, S., Latoschik, M.E.: Von Medienbeiträgen und Applikationen - ein interdisziplinäres Konzept zum Lehren und Lernen mit Augmented und Virtual Reality für die Hochschullehre. In: Knaus, T. und Merz, O. (hrsg.) Schnittstellen und Interfaces - Digitaler Wandel in Bildungseinrichtungen. S. 225–238. kopaed, Munich, Germany (2020).
135.
Byrka, J., Lewandowski, M., Spoerhase, J.: Approximating Node-Weighted k-MST on Planar Graphs. Theory of Computing Systems. (2020).
136.
Spangenberger, M.: Implementation and Evaluation of Content Encoding at Runtime for DASH Video Streaming, (2020).
137.
Carolin, W., Eisenmann, M., Latoschik, M.E., Grafe, S.: CoTeach – Connected Teacher Education. In: Schwaiger, M. (hrsg.) Boosting Virtual Reality in Learning. E.N.T.E.R (2020).
CoTeach develops and evaluates innovative teaching and learning contexts for student teachers and scholars. One work package couples the potential of VR with principles of intercultural learning to create tangible experiences with pedagogically responsible value
138.
Supervisor: Grigorjew, A.: Configuring Stream Aggregates with Asynchronous Traffic Shaping for Lower Latency, (2020).
In Time-Sensitive Networking, Asynchronous Traffic Shaping (ATS) provides a per-stream shaping mechanism that lowers the worst case per-hop latency impact of interfering traffic. However, due to the per-stream reshaping of traffic at every hop, serialization benefits on high-bandwidth links cannot be utilized. Aggregating multiple streams into a single shaper state reduces their bursty impact and can possibly provide better latency guarantees. This thesis is designed around the idea of aggregate configuration, i.e., when do the benefits justify the extra configuration, and how does aggregate shaping influence the subsequent hops on the streams' paths.
139.
Walter, J., Zink, J., Baumeister, J., Wolff, A.: Layered Drawing of Undirected Graphs with Generalized Port Constraints. 28th International Symposium on Graph Drawing and Network Visualization (GD 2020) (2020).
The aim of this research is a practical method to draw cable plans of complex machines. Such plans consist of electronic components and cables connecting specific ports of the components. Since the machines are configured for each client individually, cable plans need to be drawn automatically. The drawings must be well readable so that technicians can use them to debug the machines. In order to model plug sockets, we introduce port groups; within a group, ports can change their position (which we use to improve the aesthetics of the layout), but together the ports of a group must form a contiguous block. We approach the problem of drawing such cable plans by extending the well-known Sugiyama framework such that it incorporates ports and port groups. Since the framework assumes directed graphs, we propose several ways to orient the edges of the given undirected graph. We compare these methods experimentally, both on real-world data and synthetic data that carefully simulates real-world data. We measure the aesthetics of the resulting drawings by counting bends and crossings. Using these metrics, we compare our approach to Kieler [JVLC 2014], a library for drawing graphs in the presence of port constraints.
140.
Seufert, M., Wehner, N., Wieser, V., Casas, P., Capdehourat, G.: Mind the (QoE) Gap: On the Incompatibility of Web and Video QoE Models in the Wild. 16th International Conference on Network and Service Management (CNSM). , Izmir, Turkey (2020).
Education Service Providers (ESPs) have a paramount role in the digitization of education, providing reliable devices for students and teachers and high quality Internet access at schools. In this paper, a large-scale, passive, in-device Quality of Experience (QoE) monitoring system is presented, which was deployed into a nationwide network of education-purpose devices. Four months' worth of continuous measurements were conducted by an ESP, covering more than 800 education centers and about 4000 devices, used both in schools and at home. When analyzing the QoE of web sessions in school networks, we identify a fundamental issue with the compatibility of web browsing and video QoE models, which inhibits the successful application of QoE-aware network management for multiple services.
141.
Glémarec, Y., Lugrin, J.-L., Bosser, A.-G., Cagniat, P., Buche, C., Latoschik, M.E.: Pushing Out the Classroom Walls: A Scalability Benchmark for a Virtual Audience Behaviour Model in Virtual Reality. (2020).
142.
Roth, D., Latoschik, M.E.: Construction of the Virtual Embodiment Questionnaire (VEQ). IEEE Transactions on Visualization and Computer Graphics. 1–1 (2020).
User embodiment is important for many virtual reality (VR) applications, for example, in the context of social interaction, therapy, training, or entertainment. However, there is no data-driven and validated instrument to empirically measure the perceptual aspects of embodiment, necessary to reliably evaluate this important phenomenon. To provide a method to assess components of virtual embodiment in a reliable and consistent fashion, we constructed a Virtual Embodiment Questionnaire (VEQ). We reviewed previous literature to identify applicable constructs and questionnaire items, and performed a confirmatory factor analysis (CFA) on the data from three experiments (N = 196). The analysis confirmed three factors: (1) ownership of a virtual body, (2) agency over a virtual body, and (3) the perceived change in the body schema. A fourth study (N = 22) was conducted to confirm the reliability and validity of the scale, by investigating the impacts of latency and latency jitter present in the simulation. We present the proposed scale and study results and discuss resulting implications.
143.
Lindner, S., Latoschik, M.-E., Rittner, H.: Virtual Reality als Baustein in der Behandlung akuter und chronischer Schmerzen. AINS-Anästhesiologietextperiodcentered Intensivmedizintextperiodcentered Notfallmedizintextperiodcentered Schmerztherapie. 55, 549–561 (2020).
Schmerzbehandlung zählt zu den täglichen Routinen klinischer Anästhesisten. Im Rahmen eines wohlüberlegten Einsatzes von Schmerzmedikamenten sind Alternativen zur medikamentösen Schmerztherapie notwendig. Virtual Reality (VR) konnte sich in den letzten Jahren durch immer kostengünstigere und bessere Technologien als realistische Ergänzung etablieren. Möglichkeiten der VR sowie Indikationen und Kontraindikationen werden aufgezeigt.
144.
Wassermann, S., Casas, P., Ben Houidi, Z., Huet, A., Seufert, M., Wehner, N., Schüler, J., Cai, S.-M., Shi, H., Xu, J., Hoßfeld, T., Rossi, D.: Are you on Mobile or Desktop? On the Impact of End-User Device on Web QoE Inference from Encrypted Traffic. 16th International Conference on Network and Service Management (CNSM). , Izmir, Turkey (2020).
Web browsing is one of the key applications of the Internet, if not the most important one. We address the problem of Web Quality of Experience (QoE) monitoring from the ISP perspective, relying on in-network, passive measurements. As a proxy to Web QoE, we focus on the analysis of the well-known SpeedIndex (SI) metric. Given the lack of app-level data visibility introduced by the wide adoption of end-to-end encryption, we resort to machine-learning models to infer the SI and the QoE level of individual web page loading sessions, using as input only packet- and flow-level data. Our study targets the impact of different end-user device types (e.g., smartphone, desktop, tablet) on the performance of such models. Empirical evaluations on a large, multi-device, heterogeneous corpus of Web QoE measurements for top popular websites demonstrate that the proposed solution can infer the SI as well as estimate QoE ranges with high accuracy, using either packet-level or flow-level measurements. In addition, we show that the device type adds a strong bias in the feasibility of these Web QoE models, putting into question the applicability of previously conceived approaches on single-device measurements. To improve the state of affairs, we conceive cross-device generalizable models operating at both packet and flow levels, offering a feasible solution for Web QoE monitoring in operational, multi-device networks. To the best of our knowledge, this is the first study tackling the analysis of Web QoE from encrypted network traffic in multi-device scenarios.
145.
Dworschak, N.-D.: Evaluating Temporal Impairments in Music Streaming, (2020).