Über die Zukunft von Sprachassistenten und die Herausforderungen neuer Technologien sowie das Projekt "MOTIV" sprachen Professorin Carolin Wienrich und Professor Andreas Hotho mit dem Bayerische Forschungsinstitut für Digitale Transformation (bidt).
moreNews - Data Science Chair
In our work we review existing approaches that explain AI-based anomaly detectors by highlighting features relevant for their predictions.
more
Carolin Wienrich, Marc Latoschik, and Andreas Hotho visited the Bavarian Digital Summit. In the evening, they were invited to the reception of State Minister Judith Gerlach.
moreThe MAGNET4Cardiac7T project will be funded.
11/15/2022
The final decision is out! The MAGNET4Cardiac7T project will be funded. The official start is on the 01.12.2022. Within the project a method for modelling the distribution of electromagnetic fields in a human thorax while using a MRT-Scanner will be developed.
more
In this paper by M. Steininger et al., deep learning helps to improve climate models by post-processing their outputs.
more
In this paper, we combine NLP with Graph Learning to advance computational literary studies - using Tolkien's Legendarium as a case study.
more
In this paper by K. Kobs, M. Steininger, and A. Hotho, we use language to guide an image embedding process such that the resulting embedding space is focused on a desired similarity notion.
moreOne Petabyte of Cluster Storage
09/30/2022The Data Science Chair at the University of Würzburg celebrates its data storage infrastructure finally reaching the milestone of 1 PetaByte raw storage capacity.
With further expansion on the horizon for later this year, our Ceph based cluster is getting ready for our future scientific endeavours that are driven by large-scale datasets from various domains, like Natural Language Processing, Environmental Research, Recommender Systems and others.
More information about our cluster can be found at the following link:
more
In this paper, we investigate if Deep Metric Learning models are prone to background bias and test a method to alleviate such bias.
moreOur paper "Towards Explainable Occupational Fraud Detection" has been accepted at MIDAS 2022
07/22/2022
In our paper, we investigate the performance and interpretability of machine learning approaches when detecting occupational fraud in company data, finding models that give both strong performance and comprehensible decisions.
more
