AdaPath: Query-Adaptive Path-Finding via Path-Bank for Multi-Hop Implicit Biomedical KGQA
2026-08-31 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors address challenges in answering complex biomedical questions using knowledge graphs, where questions often lack clear steps and the graphs are very dense. They introduce AdaPath, a system that learns special ‘meta-paths’ tailored to each question to help find the right connections in the graph. They also create a new benchmark called BioStrat-QA to test how well methods handle different types of multi-step questions. Their results show AdaPath works better than existing methods, especially when questions don’t clearly show intermediate information.
Knowledge GraphPath-findingBiomedical Question AnsweringMulti-hop ReasoningMeta-pathBioStrat-QAQuery SemanticsDense GraphKnowledge Graph Reasoning
Authors
Jun Hyeong Kim, Dongki Kim, Yinhua Piao, Sung Ju Hwang
Abstract
Path-finding over knowledge graphs has become an effective way to ground LLM reasoning on multi-hop questions. However, biomedical QA introduces two distinct challenges that general-domain methods are not designed for: (i) queries do not expose intermediate reasoning and can be answered through multiple valid pathways, and (ii) biomedical knowledge graphs are densely connected, so path-finding methods easily take wrong turns. To address these challenges, we propose AdaPath, a path-finding framework that retrieves query-adaptive meta-paths from Path-Bank, which captures both query semantics and biomedical knowledge graph structure. AdaPath provides the missing cues in biomedical queries while effectively pruning dense knowledge graph neighborhoods during multi-hop reasoning. We further release BioStrat-QA, a biomedical KGQA benchmark that stratifies multi-hop queries by how much intermediate reasoning they expose. Across biomedical KGQA benchmarks, AdaPath consistently outperforms baselines, sustaining meaningful path-finding even when multi-hop queries expose less surface information. The source code is available at https://github.com/Jun-Hyeong-Kim/AdaPath.