Enrich-on-Graph: Query-Graph Alignment for Complex Reasoning with LLM Enriching
Enrich-on-Graph (EoG) addresses knowledge gap issues in KGQA by using Large Language Models to improve Knowledge Graphs. It creates a query-aligned graph by maximizing mutual information between queries and graphs, leading to better reasoning. The three stages—Parsing, Pruning, and Enriching—align semantics efficiently while keeping computational costs low.