Saved in:
Bibliographic Details
Main Authors: Wang, Xiaochen, Wu, Zongyu, Zhong, Yuan, Zhang, Xiang, Wang, Suhang, Ma, Fenglong
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.00261
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915319695540224
author Wang, Xiaochen
Wu, Zongyu
Zhong, Yuan
Zhang, Xiang
Wang, Suhang
Ma, Fenglong
author_facet Wang, Xiaochen
Wu, Zongyu
Zhong, Yuan
Zhang, Xiang
Wang, Suhang
Ma, Fenglong
contents Graph retrieval-augmented generation (GRAG) places high demands on graph-specific retrievers. However, existing retrievers often rely on language models pretrained on plain text, limiting their effectiveness due to domain misalignment and structure ignorance. To address these challenges, we propose GPR, a graph-based retriever pretrained directly on knowledge graphs. GPR aligns natural language questions with relevant subgraphs through LLM-guided graph augmentation and employs a structure-aware objective to learn fine-grained retrieval strategies. Experiments on two datasets, three LLM backbones, and five baselines show that GPR consistently improves both retrieval quality and downstream generation, demonstrating its effectiveness as a robust retrieval solution for GRAG.
format Preprint
id arxiv_https___arxiv_org_abs_2506_00261
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GPR: Empowering Generation with Graph-Pretrained Retriever
Wang, Xiaochen
Wu, Zongyu
Zhong, Yuan
Zhang, Xiang
Wang, Suhang
Ma, Fenglong
Information Retrieval
Computation and Language
Graph retrieval-augmented generation (GRAG) places high demands on graph-specific retrievers. However, existing retrievers often rely on language models pretrained on plain text, limiting their effectiveness due to domain misalignment and structure ignorance. To address these challenges, we propose GPR, a graph-based retriever pretrained directly on knowledge graphs. GPR aligns natural language questions with relevant subgraphs through LLM-guided graph augmentation and employs a structure-aware objective to learn fine-grained retrieval strategies. Experiments on two datasets, three LLM backbones, and five baselines show that GPR consistently improves both retrieval quality and downstream generation, demonstrating its effectiveness as a robust retrieval solution for GRAG.
title GPR: Empowering Generation with Graph-Pretrained Retriever
topic Information Retrieval
Computation and Language
url https://arxiv.org/abs/2506.00261