AMIA 2025 NLP Annual Symposium - Lightning talk

We are excited to announce that our paper, “Efficient and Scalable Retrieval in Knowledge Graphs: From Single-Graph to Partitioned Multi-Graph Approaches,” has been accepted to the AMIA 2025 NLP Annual Symposium!

In this paper, we propose an efficient and scalable framework for multi-hop retrieval in knowledge graphs (KGs). Our framework transforms KG traversal into sparse matrix operations and extends it to handle very large KGs through degree-aware graph partitioning, cross-graph routing, and on-demand caching.

Multi-hop traversal in KGs is a fundamental operation in many LLM+KG scenarios, including retrieval-augmented generation (RAG), knowledge verification, and reasoning. This work contributes to the integration of KGs with LLMs and enables many downstream applications.

This research is a collaboration with:

• Dr. Majid Afshar (University of Wisconsin–Madison)
• Dr. Dmitriy Dligach (Loyola University Chicago)
• Dr. Timothy Miller (Boston Children’s Hospital)

Join us for the lightning talk! Dr. He Cheng will present our work on November 16 at 11:00 AM. Slides and code will be released soon. Stay tuned!