Department of Computational and Data Sciences
Department Seminar
Speaker : Prof. Michael Färber, Professor (W3) at the AI Center ScaDS.AI and TU Dresden, Germany
Title : From Papers to Insights with LLMs and Knowledge Graphs
Date & Time : September 01st, 2026 (Tuesday), 10:00 AM
Venue : # 102, CDS Seminar Hall
ABSTRACT:
Science is producing knowledge faster than any human can read. The key challenge is therefore no longer access to papers, but turning them into reliable, verifiable insights. In this talk, I explain how large language models and knowledge graphs can work together to extract claims and evidence from the literature, connect information at scale, and generate answers that remain traceable to sources. I also discuss “AI with a conscience” and present practical methods for controlling model behavior through prompting and constrained inference. In addition, I outline mechanisms that flag uncertainty and potentially harmful outputs and enable knowledge updates without full retraining. The goal is to build AI systems that support scientists while remaining transparent, controllable, and aligned with human values.
BIOGRAPHY:
Prof. Dr.-Ing. Michael Färber (https://faerber-lab.github.io/) is a Full Professor (W3) at the AI Center ScaDS.AI and TU Dresden, Germany, where he leads the “Scalable Software Architectures for Data Analytics” group. He previously served as Deputy Full Professor for Web Science at the Karlsruhe Institute of Technology (KIT). His research focuses on large language models, knowledge graphs, and graph neural networks, with an emphasis on trustworthy AI for science. He develops systems for large-scale analysis of scientific literature and evidence-grounded text generation with verifiable citations. He has published 130+ peer-reviewed papers at venues such as AAAI, ACL, EMNLP, CIKM, KDD, NAACL, and ICML.
Host Faculty: Dr. Danish Pruthi
ALL ARE WELCOME
