Photo credit: Anna Logue

DWS Colloquium HWS2020

The DWS colloquium takes place every Tuesday from 13:45 to 15:15.  The colloquium consists of a series of scientific talks as well as several poster sessions in which the DWS PhD students present their research.

The colloquium is a joint event of all DWS senior faculty members (Prof. Stuckenschmidt, Prof. Bizer, Prof. Paulheim, Prof. Ponzetto and Prof. Gemulla). 

Due to the Corona-situation, the scientific talks of the HWS2020 colloquium will take place via Zoom (virtual room: WIM-ZOOM-4). The poster session will talk place via MS Teams where the posters will be presented in parallel in different MS Teams chanels.

Colloquium Schedule




29.09.2020Heiko Paulheim Filter Bubbles and Algorithmic Pricing
Han van der Aa Process Analytics at DWS: A Research Agenda
6.10.2020Dr. Michael Roth
(University of Stuttgart)
Understanding Misunderstandings Through Revisions of WikiHow Articles
Christian Bartelt Post-hoc Interpretation of Artifical Neural Networks by Parameter Processing

Poster Session 1
Amirhossein Kardoost
Adrian Rebmann
Anna Primpeli
Anne Lauscher
Jovita Lukasik
Jan Portisch
Kalun Ho

Amirhossein: Self-supervised sparse to dense motion segmentation
Adrian: Extracting Semantic Process Information from Event Logs
Anna: Active Learning for Multi Source Entity Resolution
Anne: On the Limitations of Zero-Shot Cross-Lingual Transfer with Multilingual Transformers
Jovita: Smooth Variational Graph Embeddings for Efficient Neural Architecture Search
Jan: Supervised and Embedding-Based Schema Matching
Kalun: A Multi-Tier Approach for Scalable Image Clustering

20.10.2020Dr. Pasquale Minervini
(University College London)
Explainable, Verifiable, and Data-Efficient Deep Learning
Dr. Sebastijan Dumancic
(KU Leuven)
DeepProbLog: Neural Probabilistic Logic Programming
27.10.2020Ralph PeetersDual-Objective Fine-tuning and Explaining BERT for Product Matching
Jonas Pfeiffer (Universität Darmstadt)Adapters in Transformers - A New Paradigm for Transfer Learning…?
3.11.2020Edoardo Ponti (University of Cambridge)Sample Efficiency and Generalisation in Multilingual NLP
Mario Fritz (CISPA Saarbrücken)Trustworthy Machine Learning & Privacy
10.11.2020Poster Session 2
Alexander Brinkmann
Bernhard Schäfer
Kilian Theil
David Friede
Jonathan Kobbe
Nicolas Heist
Robert Litschko
Sven Hertling
Alexander: Hierarchical Product Classification using Transformers
Bernhard: Handwritten Business Process Recognition
Kilian: Experimental Evalution of Uncertainty Perception in Investment Decisions
David: Teaching Transformers to Reason through Weak Supervision

Jonathan: Stance Detection for Arguments
Nicolas: Novel Entity Disambiguation in Wikipedia Listings
Robert: Evaluating Multilingual Text Encoders for Unsupervised Cross-Lingual Retrieval

Sven: Multi Knowledge Graph Matching
17.11.2020Olaf Hartig (Linköping University)

The RDF* and SPARQL* Approach to Annotate Statements in RDF and to Achieve Interoperability of Graph Database Systems

Daniel RuffinelliIn Search of Holistic Representations of Knowledge Graphs
24.11.2020Emile van Krieken (VU Amsterdam)Analyzing Differentiable Fuzzy Logic Operators
Adrian KochsiekScaling Knowledge Graph Embeddings to Large-scale Graphs
1.12.2020Anna Primpeli (at) 3:15 pmMulti-Source Entity Matching with Active Learning
David Jurgens (University of Michigan) (at) 4:00 pmSocial Interpretations of Interpersonal Communication
8.12.2020Matthias Weidlich (HU Berlin) (at) 4:15 pmState-management in Complex Event Processing
Vered Shwartz (University of Washington) (at) 5:00 pmUnsupervised Methods for Commonsense Reasoning

Poster Session 3 (at) 1:45 pm
Andreea Iana
Alexander Kraus
Chia-Chien Hung
Diana Sola
Fabian David Schmidt
Michael Schlechtinger
Ralph Peeters
Steffen Jung

Andreea: „Terroristen, Mörder, und Vergewaltiger - Das Klientel der Grünen?“: Knowledge Graph Construction for (Unbiased?) News Recommendation
Alexander: System-state analysis of business processes
Chia-Chien: Bilingual Representations for Rare Words with Adversarial Training
Diana: A Rule-based Recommendation Approach for Business Process Modeling
Fabian: Lexically-constrained cross-lingual representation learning
Michael: Hidden Cooperation of Pricing Algorithms
Ralph: Explaining (Dual-Objective) BERT for Product Matching
Steffen: Spectral Distribution aware Image Generation

12.1.2021Steffen JungApproaching Correlation Clustering with Graph Convolutional Neural Networks
19.1.2021Tobias WellerIntroduction to the Data and Web Science Group
Alexander BrinkmannHierarchical Product Classification using Transformers
26.1.2021Dirk Hovy (Bocconi University) 
Patrick Betz