Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wang, Qi, Cao, Yixuan, Liu, Yifan, Zhao, Jiangtao, Luo, Ping
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2507.00477
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908646548439040
author Wang, Qi
Cao, Yixuan
Liu, Yifan
Zhao, Jiangtao
Luo, Ping
author_facet Wang, Qi
Cao, Yixuan
Liu, Yifan
Zhao, Jiangtao
Luo, Ping
contents A Retrieval-Augmented Generation (RAG)-based question-answering (QA) system enhances a large language model's knowledge by retrieving relevant documents based on user queries. Discrepancies between user queries and document phrasings often necessitate query rewriting. However, in specialized domains, the rewriter model may struggle due to limited domain-specific knowledge. To resolve this, we propose the R\&R (Read the doc before Rewriting) rewriter, which involves continual pre-training on professional documents, akin to how students prepare for open-book exams by reviewing textbooks. Additionally, it can be combined with supervised fine-tuning for improved results. Experiments on multiple datasets demonstrate that R\&R excels in professional QA across multiple domains, effectively bridging the query-document gap, while maintaining good performance in general scenarios, thus advancing the application of RAG-based QA systems in specialized fields.
format Preprint
id arxiv_https___arxiv_org_abs_2507_00477
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Read the Docs Before Rewriting: Equip Rewriter with Domain Knowledge via Continual Pre-training
Wang, Qi
Cao, Yixuan
Liu, Yifan
Zhao, Jiangtao
Luo, Ping
Information Retrieval
A Retrieval-Augmented Generation (RAG)-based question-answering (QA) system enhances a large language model's knowledge by retrieving relevant documents based on user queries. Discrepancies between user queries and document phrasings often necessitate query rewriting. However, in specialized domains, the rewriter model may struggle due to limited domain-specific knowledge. To resolve this, we propose the R\&R (Read the doc before Rewriting) rewriter, which involves continual pre-training on professional documents, akin to how students prepare for open-book exams by reviewing textbooks. Additionally, it can be combined with supervised fine-tuning for improved results. Experiments on multiple datasets demonstrate that R\&R excels in professional QA across multiple domains, effectively bridging the query-document gap, while maintaining good performance in general scenarios, thus advancing the application of RAG-based QA systems in specialized fields.
title Read the Docs Before Rewriting: Equip Rewriter with Domain Knowledge via Continual Pre-training
topic Information Retrieval
url https://arxiv.org/abs/2507.00477