dorsal/arxiv
View SchemaRelink: Constructing Query-Driven Evidence Graph On-the-Fly for GraphRAG
| Authors | Manzong Huang, Chenyang Bu, Yi He, Xingrui Zhuo, Xindong Wu |
|---|---|
| Categories | |
| ArXiv ID | 2601.07192vv1 |
| URL | https://arxiv.org/abs/2601.07192 |
| License | http://creativecommons.org/licenses/by-nc-sa/4.0/ |
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.
{
"annotation_id": "d596aa25-a656-44e0-ad56-86bed2d91ddf",
"date_created": "2026-02-17T05:53:11.755000Z",
"date_modified": "2026-02-17T05:53:11.755000Z",
"file_hash": "b97ff208e98816ed1148152b996d4422f3b32711c79f4b7e7a54045f14b1495a",
"private": false,
"record": {
"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\u0027s inherent incompleteness often breaks reasoning paths. Second, the graph\u0027s low signal-to-noise ratio introduces distractor facts, presenting query-relevant but misleading knowledge that disrupts the reasoning process.\n 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.\n 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.",
"arxiv_id": "2601.07192",
"authors": [
"Manzong Huang",
"Chenyang Bu",
"Yi He",
"Xingrui Zhuo",
"Xindong Wu"
],
"categories": [
"cs.CL",
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
"title": "Relink: Constructing Query-Driven Evidence Graph On-the-Fly for GraphRAG",
"url": "https://arxiv.org/abs/2601.07192",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "fdfe8a23-ee06-4695-bef9-33572046ced6",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}