Katja Hose

Professor at TU Wien’s Databases and Artificial Intelligence research unit

Katja Hose is a full professor at TU Wien, where she leads the DMKI Lab (Data Management and Knowledge-Driven AI). She previously held positions at Aalborg University and the Max Planck Institute for Informatics, and received her PhD from Ilmenau University of Technology. Her research is rooted in data management and knowledge engineering, with a focus on graph databases, knowledge graphs, data integration, and applied machine learning. She has co-authored more than 150 peer-reviewed publications and serves regularly as a reviewer for major conferences including TheWebConf/WWW, VLDB, SIGMOD, EDBT, ISWC, and ESWC. She sits on the editorial boards of the VLDB Journal and the Semantic Web Journal, and serves as an associate editor for TGDK. Katja has also held numerous organizational roles for international conferences, including program co-chair positions for VLDB 2027, ISWC 2024, EDBT 2023, and ESWC 2021.

https://www.dbai.tuwien.ac.at/staff/khose/About_me.html

Title: Reliable Knowledge in the Age of Generative AI: From Noisy Data to Trustworthy Agents

Generative AI is reshaping the way information is produced, accessed, and acted upon, yet its effectiveness ultimately depends on the quality and reliability of the knowledge it draws from. As these systems grow more capable, the need for structured, trustworthy, and well-understood data becomes increasingly central – not only for ensuring factual correctness but also for enabling meaningful interactions between symbolic and statistical methods. This talk reflects on the role of knowledge graphs in supporting reliable AI, highlighting how data quality can be strengthened at scale and why such foundations matter for hybrid approaches that combine knowledge-driven reasoning with generative models. It also touches on recent insights into hallucinations and emerging directions involving agentic systems.