To the OTH Regensburg website

Teaching

Research and teaching are closely intertwined: the latest findings from the projects are incorporated directly into students’ education. Furthermore, teaching itself becomes a subject of research. The ADKI project is developing AI-based tutoring systems that provide learners with interactive and personalised support, for example by analysing their attempts to solve programming tasks using tests or formal verification tools, and deriving targeted guidance from this analysis.

An overview of our courses:

  • Algorithms and Data Structures
  • Theoretical Computer Science
  • Logic and Knowledge Representation
  • Explainable AI
  • Logical and Functional Programming
  • Databases
  • Programming 2
  • Operating Systems

Student Projects and Theses

Interested students can get directly involved in research through Bachelor’s and Master’s theses, research placements, Research Master’s projects or as student assistants. The topics advertised are intended as a starting point and can be adapted to individual interests; students are expressly encouraged to put forward their own proposals. Current opportunities are published on the laboratory’s Moodle course.

  • Counterfactual explanations for GraphRAG systems
  • Graph embeddings for regulatory knowledge graphs
  • Evidence-based data storytelling in auditing
  • Knowledge graph engineering for domain knowledge
  • LLM tutor with tool-assisted diagnosis
  • Price regression and similarity search for manufactured parts (with P&S Maschinenbau)

Research

The focus is on combining data-driven AI methods with formal knowledge to develop transparent and trustworthy systems. Research is being conducted into how large language models, knowledge graphs and ontologies can be integrated with logic-based methods, for example to automatically transform complex data analyses into comprehensible data stories. Other key areas of focus include agentic coding, in which AI agents use formalised rules to develop software in a traceable manner, and the efficient evaluation of database queries across constantly changing data.

  • Data Tales (BayVFP): Data storytelling with hybrid AI for auditing
  • KIGA (BayVFP): AI-supported business process analysis through hybrid process mining – that is, the combination of interactive ML methods with knowledge-based systems – in collaboration with dab GmbH and the University of Bamberg. Prof. Scheele was the project leader. Results include, amongst other things, semantically enriched event logs, explainable clustering methods and an LLM-based explanation module
  • ADKI: AI-based tutoring systems, awarded the OTH Prize for Innovation and Quality in Teaching
  • F-IVM: Analyses and learning methods for relational databases undergoing continuous updates, a research project involving Prof. Kara (Universities of Oxford, Edinburgh and Zurich)

Contact

Staff