Relink: Constructing Query-Driven Evidence Graph On-the-Fly for GraphRAG

TL;DR

Relink introduces a dynamic, query-specific evidence graph construction method for GraphRAG, improving reasoning accuracy by addressing knowledge graph incompleteness and distractor facts, leading to significant performance gains in open-domain question answering.

Contribution

It proposes a novel reason-and-construct paradigm with Relink, which dynamically builds evidence graphs tailored to each query, enhancing reasoning and accuracy over static KG-based methods.

Findings

Achieves 5.4% higher EM and 5.2% higher F1 scores on five QA benchmarks.

Effectively repairs incomplete knowledge graphs by instantiating facts from a latent relation pool.

Reduces distractor influence by jointly evaluating KG and latent relations for query relevance.

Abstract

Graph-based Retrieval-Augmented Generation (GraphRAG) mitigates hallucinations in Large Language Models (LLMs) by grounding them in structured knowledge. However, current GraphRAG methods are constrained by a prevailing \textit{build-then-reason} paradigm, which relies on a static, pre-constructed Knowledge Graph (KG). This paradigm faces two critical challenges. First, the KG's inherent incompleteness often breaks reasoning paths. Second, the graph's low signal-to-noise ratio introduces distractor facts, presenting query-relevant but misleading knowledge that disrupts the reasoning process. To address these challenges, we argue for a \textit{reason-and-construct} paradigm and propose Relink, a framework that dynamically builds a query-specific evidence graph. To tackle incompleteness, \textbf{Relink} instantiates required facts from a latent relation pool derived from the original text corpus, repairing broken paths on the fly. To handle misleading or distractor facts, Relink employs a unified, query-aware evaluation strategy that jointly considers candidates from both the KG and latent relations, selecting those most useful for answering the query rather than relying on their pre-existence. This empowers Relink to actively discard distractor facts and construct the most faithful and precise evidence path for each query. Extensive experiments on five Open-Domain Question Answering benchmarks show that Relink achieves significant average improvements of 5.4\% in EM and 5.2\% in F1 over leading GraphRAG baselines, demonstrating the superiority of our proposed framework.