Efficient Document Retrieval with G-Retriever

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Solanki, Manthankumar
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910915478159360
author Solanki, Manthankumar
author_facet Solanki, Manthankumar
contents Textual data question answering has gained significant attention due to its growing applicability. Recently, a novel approach leveraging the Retrieval-Augmented Generation (RAG) method was introduced, utilizing the Prize-Collecting Steiner Tree (PCST) optimization for sub-graph construction. However, this method focused solely on node attributes, leading to incomplete contextual understanding. In this paper, we propose an enhanced approach that replaces the PCST method with an attention-based sub-graph construction technique, enabling more efficient and context-aware retrieval. Additionally, we encode both node and edge attributes, leading to richer graph representations. Our method also incorporates an improved projection layer and multi-head attention pooling for better alignment with Large Language Models (LLMs). Experimental evaluations on the WebQSP dataset demonstrate that our approach is competitive and achieves marginally better results compared to the original method, underscoring its potential for more accurate question answering.
format Preprint
id arxiv_https___arxiv_org_abs_2504_14955
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Efficient Document Retrieval with G-Retriever
Solanki, Manthankumar
Machine Learning
Textual data question answering has gained significant attention due to its growing applicability. Recently, a novel approach leveraging the Retrieval-Augmented Generation (RAG) method was introduced, utilizing the Prize-Collecting Steiner Tree (PCST) optimization for sub-graph construction. However, this method focused solely on node attributes, leading to incomplete contextual understanding. In this paper, we propose an enhanced approach that replaces the PCST method with an attention-based sub-graph construction technique, enabling more efficient and context-aware retrieval. Additionally, we encode both node and edge attributes, leading to richer graph representations. Our method also incorporates an improved projection layer and multi-head attention pooling for better alignment with Large Language Models (LLMs). Experimental evaluations on the WebQSP dataset demonstrate that our approach is competitive and achieves marginally better results compared to the original method, underscoring its potential for more accurate question answering.
title Efficient Document Retrieval with G-Retriever
topic Machine Learning
url https://arxiv.org/abs/2504.14955