[2603.22633] Graph-Aware Late Chunking for Retrieval-Augmented Generation in Biomedical Literature

Computer Science > Artificial Intelligence arXiv:2603.22633 (cs) [Submitted on 23 Mar 2026] Title:Graph-Aware Late Chunking for Retrieval-Augmented Generation in Biomedical Literature Authors:Pouria Mortezaagha, Arya Rahgozar View a PDF of the paper titled Graph-Aware Late Chunking for Retrieval-Augmented Generation in Biomedical Literature, by Pouria Mortezaagha and 1 other authors View PDF HTML (experimental) Abstract:Retrieval-Augmented Generation (RAG) systems for biomedical literature are typically evaluated using ranking metrics like Mean Reciprocal Rank (MRR), which measure how well the system identifies the single most relevant chunk. We argue that for full-text scientific documents, this paradigm is incomplete: it rewards retrieval precision while ignoring retrieval breadth -- the ability to surface evidence from across a document's structural sections. We propose GraLC-RAG, a framework that unifies late chunking with graph-aware structural intelligence, introducing structure-aware chunk boundary detection, UMLS knowledge graph infusion, and graph-guided hybrid retrieval. We evaluate six strategies on 2,359 IMRaD-filtered PubMed Central articles using 2,033 cross-section questions and two metric families: standard ranking metrics (MRR, Recall@k) and structural coverage metrics (SecCov@k, CS Recall). Our results expose a sharp divergence: content-similarity methods achieve the highest MRR (0.517) but always retrieve from a single section, while structure-aware meth...