ComRAG: Retrieval-Augmented Generation with Dynamic Vector Stores for Real-time Community Question Answering in Industry

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Qinwen, Tao, Wenbiao, Zhu, Zhiwei, Xi, Mingfan, Guo, Liangzhong, Wang, Yuan, Wang, Wei, Lan, Yunshi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915365929353216
author Chen, Qinwen
Tao, Wenbiao
Zhu, Zhiwei
Xi, Mingfan
Guo, Liangzhong
Wang, Yuan
Wang, Wei
Lan, Yunshi
author_facet Chen, Qinwen
Tao, Wenbiao
Zhu, Zhiwei
Xi, Mingfan
Guo, Liangzhong
Wang, Yuan
Wang, Wei
Lan, Yunshi
contents Community Question Answering (CQA) platforms can be deemed as important knowledge bases in community, but effectively leveraging historical interactions and domain knowledge in real-time remains a challenge. Existing methods often underutilize external knowledge, fail to incorporate dynamic historical QA context, or lack memory mechanisms suited for industrial deployment. We propose ComRAG, a retrieval-augmented generation framework for real-time industrial CQA that integrates static knowledge with dynamic historical QA pairs via a centroid-based memory mechanism designed for retrieval, generation, and efficient storage. Evaluated on three industrial CQA datasets, ComRAG consistently outperforms all baselines--achieving up to 25.9% improvement in vector similarity, reducing latency by 8.7% to 23.3%, and lowering chunk growth from 20.23% to 2.06% over iterations.
format Preprint
id arxiv_https___arxiv_org_abs_2506_21098
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ComRAG: Retrieval-Augmented Generation with Dynamic Vector Stores for Real-time Community Question Answering in Industry
Chen, Qinwen
Tao, Wenbiao
Zhu, Zhiwei
Xi, Mingfan
Guo, Liangzhong
Wang, Yuan
Wang, Wei
Lan, Yunshi
Computation and Language
Artificial Intelligence
Community Question Answering (CQA) platforms can be deemed as important knowledge bases in community, but effectively leveraging historical interactions and domain knowledge in real-time remains a challenge. Existing methods often underutilize external knowledge, fail to incorporate dynamic historical QA context, or lack memory mechanisms suited for industrial deployment. We propose ComRAG, a retrieval-augmented generation framework for real-time industrial CQA that integrates static knowledge with dynamic historical QA pairs via a centroid-based memory mechanism designed for retrieval, generation, and efficient storage. Evaluated on three industrial CQA datasets, ComRAG consistently outperforms all baselines--achieving up to 25.9% improvement in vector similarity, reducing latency by 8.7% to 23.3%, and lowering chunk growth from 20.23% to 2.06% over iterations.
title ComRAG: Retrieval-Augmented Generation with Dynamic Vector Stores for Real-time Community Question Answering in Industry
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2506.21098