Seminar on Knowledge Graphs and Large Language Models (FSS 2024)

Large Language Models (LLMs) like ChatGPT have been among the most omnipresent advances in the field of Artificial Intelligence in the past years. Despite their potential and versatility, they have also exposed some drawbacks, such as hallucinations, intransparency, and non-factuality. Knowledge Graphs (KGs), on the hand, are a main driver of white box AI, which are known to be transparent and factual.

In this seminar, we will look at works at the cross-roads between knowledge graphs and large language models, which aim at combining the best of both worlds. Areas of interest include, but are not limited to:

  • KGs for LLM training and tuning
  • KG Enhanced LLM inference and explanations
  • Prompt engineering with KGs
  • Completing/Refining KGs with LLMs
  • Question Answering/Text Generation with KGs and LLMs
  • Joint Reasoning with KGs and LLMs

Goals

In this seminar, you will familiarize yourself with approaches to knowledge graph construction. You will read research papers, specifications, and tool descriptions, as well as conduct own experiments where applicable, and you will discuss the insights with the other participants of the seminar.

As a participant, you are supposed to introduce a particular technique for knowledge graph construction and present it to the seminar participants. Each seminar paper undergoes a peer review process in the seminar. Presentations are supposed to be about 25 minutes long.

Organization

This seminar is organized by Prof. Dr. Heiko Paulheim

Available for Master students (2 SWS, 4 ECTS)

Prerequisites: Basic prior knowledge in knowledge graphs (e.g., by attending IE650) and LLMs

Additional resources:

Schedule

  • February 22nd, 13:45–15:15: kick off meeting, room tba
  • March 4th: topic assignment
  • March 31st: paper draft due
  • April 14th: peer review due
  • May 2nd, 16th, 23rd: 13:45–17:00 presentations and discussions, schedule and room tba
  • June 16th: final paper due