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

2026-01-12

arXiv.org

AI Summary

Existing GraphRAG approaches are constrained by static knowledge graphs, which often suffer from incomplete structures that disrupt reasoning paths and are further compromised by low signal-to-noise ratios in factual evidence. To address these limitations, this work proposes Relink, a novel framework that introduces a “reason-and-construct” paradigm to dynamically build query-oriented evidence graphs. Relink instantiates missing relations on-the-fly from raw text, synergistically integrating structured knowledge graphs with a latent relation pool, and employs a query-aware unified scoring mechanism to jointly select high-quality candidate facts. This approach adaptively repairs broken reasoning chains and proactively filters noise, substantially enhancing the faithfulness and precision of the resulting evidence graph. Evaluated on five open-domain question answering benchmarks, Relink achieves consistent improvements, averaging +5.4% in Exact Match and +5.2% in F1 score over state-of-the-art GraphRAG baselines.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Reasoning with Beliefs
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

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.

Problem

• GraphRAG
• knowledge graph incompleteness
• distractor facts
• reasoning path
• signal-to-noise ratio

Innovation

• GraphRAG
• dynamic evidence graph
• query-aware reasoning
• latent relation extraction
• reason-and-construct paradigm