Franz Krause Defended his PhD

In his thesis, which was largely an outcome of the EU-funded Teaming.AI project, Franz has worked on topics such as

1. The modeling of BPMN processes and process trails in OWL and RDF, leading to a formalism that allows semantic analyses over processes and process executions incorporating domain knowledge.
2. Embeddings for dynamic knowledge graphs using local reconstructions, combining ideas from knowledge graph embedding and graph neural networks, and creating a novel knowledge graph embedding mechanism solely based on reconstruction.
3. A method for explaining predictions made by Graph Neural Networks on Knowledge Graphs, which are usually black box models not allowing for any interpretation.

The evaluation committee, consisting of Bernhard Moser (SCC Hagenberg and JKU Linz), Daniel Schuster (University of Mannheim), Heiner Stuckenschmidt (University of Mannheim) and Heiko Paulheim (University of Mannheim), unanimously recommended to pass his dissertation. Congratulations, Franz!