Thomas Hoppe

After a short intermezzo in the automotive industry and his doctoral thesis in logic programming, Thomas started his research on search engines in several divisions of Deutsche Telekom. He founded a startup on semantic search technology with three partners as CEO, worked in a small company and he was also a freelancing data scientist.
In 2018 he joined Fraunhofer FOKUS as researcher in the data analytics group, and led several projects about information and knowledge extraction, semantic search and modelling supply chains as knowledge graph.
In the GenAI4Media-Projekt he investigates approaches for confabulation detection of LLMs, knowledge graph extraction from news, knowledge injection into the attention mechanism of LLMs. Recently he identified a new approach for the similarity preserving layout of knowledge graphs.
Between 2015 and 2026 he was lecturer and appointed as guest professor at the HTW-Berlin (university of applied science) for databases, information systems and data science. He is editor and author of several articles and books about applied AI and Semantic Search, and leading co-organizer of the annual workshop series “Applied Machine Intelligence”.