In May, the new research project “SynDIKAT” was launched. The acronym stands for “Synthetic Disinformation Data, Infrastructure for Collaboration and Analysis Toolbox for Social Media Research”. Funded for three years by the German Federal Ministry of Research, Technology and Space (BMFTR), the collaborative project focuses on disinformation in social media.

The project develops approaches for the privacy-compliant sharing and generation of synthetic social media data using large language models (LLMs) and agent-based simulations. Its main focus is on methodological questions in disinformation research.

According to the researchers, one central problem is the lack of shared social media datasets. As a result, methods for detecting and combating disinformation can currently only be compared and evaluated to a limited extent. SynDIKAT therefore aims to enable legally compliant data sharing and improve the evaluation of relevant algorithms for pattern and anomaly detection.

The project is led by Prof. Dr. Christian Grimme, Janina Lütke-Stockdiek, and Michael Vogt from the Department of Information Systems, as well as Prof. Dr. Thorsten Quandt, Malin Zoe Richter, and Paula Philine Jung from the Department of Communication at the University of Münster. Additional partners include GESIS – Leibniz Institute for the Social Sciences, the Center for Advanced Internet Studies, and TU Dortmund University.

Links related to this news
BMFTR SynDIKAT

Press release from the University of Münster