Diego Collarana-Vargas

Fraunhofer Institute for Intelligent Analysis and Information Systems, IAIS, Germany

Team Leader Knowledge Graphs


Dr. Diego Collarana (male) is a Senior Research Engineer and Team Leader at Fraunhofer IAIS. Dr. Collarana has more than 15 years of experience conceptualizing and developing enterprise information systems using state-of-the-art AI technologies, i.e., Knowledge Graphs and Deep Learning. His master’s studies (Double degree from Universidad Politecnica de Madrid and Technische Universität Kaiserslautern) in Software Engineering gives him a solid knowledge of Project Management, Software Architecture, Design, and Development. Furthermore, in 2018, he obtained a Ph.D. (Magna Cum Laude) at the University of Bonn on Artificial Intelligence. Dr. Collarana’s experience allows him to solve complex enterprise problems, including Knowledge Graphs construction, GNNs, NLU, and Conversational AI. He is the founder/leader of several open-source projects, including MINTE, a semantic integration technique that allows companies to go from heterogeneous data to actionable knowledge. The MINTE technique implements a set of entity similarities techniques that work on noisy data, i.e., data with interoperability conflicts. Dr. Collarana authored more than 40 articles in international journals and conferences and 2 best paper awards. He served as a Program Committee member of more than 20 conferences.

Talks and Events

Knowledge Graph Treatments for Hallucinating Large Language Models

Despite the excitement about Large Language Models (LLM), these models suffer from hallucinations problems, e.g., generating factually incorrect text. These problems restrict the development of production-ready applications. This talk will highlight the importance of combining Knowledge Graphs with Large Language Models to develop industry-ready applications. We will present different approaches, from pragmatic to under-research approaches to threat and handle hallucination problems using Knowledge Graphs at different phases of the LLM lifecycle. We will accompany our presentation with use cases that Fraunhofer works with partners from big German industries under the OpenGPT-X project.

Track: NLP

Session Topics:

  • Natural Language Processing and Understanding
  • Conversational Interfaces
  • Deep Learning for and with Knowledge Graphs