Empowering Large Language Models to Set Up Knowledge Retrieval Indexing via Self-Learning

计算机科学 搜索引擎索引 集合(抽象数据类型) 情报检索 模式(遗传算法) 自然语言处理 人工智能 语言模型 基于知识的系统 领域知识 知识检索 知识表示与推理 知识获取 知识抽取 知识组织 知识工程 编码(集合论) 自动索引 基线(sea) 知识整合 知识库
作者
Simin Niu,Mengwei Wang,Xun Liang,Zhiyu li,Sensen Zhang,Shichao Song,Hanyu Wang,Jiawei Yang,Feiyu Xiong,Chenyang Xi
出处
期刊:IEEE Transactions on Knowledge and Data Engineering [IEEE Computer Society]
卷期号:38 (3): 1710-1724 被引量:1
标识
DOI:10.1109/tkde.2025.3649907
摘要

Retrieval-augmented generation (RAG) provides an efficient solution for expanding the knowledge boundaries of large language models (LLMs), where the indexing serves as a compass to guide LLMs in locating query-relevant external knowledge. Nevertheless, current indexing methods commonly encounter a critical challenge: native indexing is convenient to construct, but it usually disrupts contextual associations and constrains the expressive capacity of rich knowledge. Conversely, knowledge indexing can structure contextual knowledge, but it is often based on preset schemas that limit its generalizability. To address it, we propose a universal and flexible knowledge indexing called pseudo-graph (PG) indexing. During the indexing construction phase, we use the advanced LLMs to transform the knowledge of each raw text into a concise and structured mind map, organizing intra-document knowledge. Subsequently, independent mind maps are linked by associating highly relevant topics or consistent facts across documents, thereby establishing inter-document knowledge connections. Eventually, using the resulting knowledge network PG as the knowledge indexing can circumvent the challenges associated with schema design reliant on preset knowledge and relationship types. During the knowledge retrieval phase, we develop a PG knowledge retriever to mimic human note-reviewing, adaptively navigating and recalling query-relevant knowledge from PG. Experimental results demonstrate that retrieving relevant pseudo-subgraphs from the PG via PG indexing and retriever significantly improves performance in fact-based Q&A, hallucination correction, and two multi-document Q&A tasks, achieving$F1_{QE}$improvements of 15.85%, 8.12%, 3.34%, and 5.73%, respectively, and outperforming the state-of-the-art baseline KGP-LLaMA. Our code is available at:https://github.com/IAAR-Shanghai/PGRAG.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
fx发布了新的文献求助10
刚刚
1秒前
he发布了新的文献求助10
1秒前
科研通AI6.4的应助被简单生活采纳,获得10
1秒前
2秒前
杜梦婷发布了新的文献求助10
2秒前
qinglan完成签到,获得积分10
2秒前
zhang5657完成签到,获得积分10
2秒前
sl发布了新的文献求助10
3秒前
李爱国的应助被醉熏的寄松采纳,获得10
4秒前
SciGPT的应助被醉熏的寄松采纳,获得10
4秒前
4秒前
Orange的应助被园园采纳,获得10
4秒前
Joany发布了新的文献求助10
4秒前
有缘发布了新的文献求助10
4秒前
huhu发布了新的文献求助10
5秒前
5秒前
孤独的文献123完成签到,获得积分10
6秒前
TRY发布了新的文献求助10
6秒前
顾矜的应助被大力的源智采纳,获得10
6秒前
6秒前
7秒前
ypp发布了新的文献求助10
8秒前
9秒前
ll发布了新的文献求助10
10秒前
10秒前
11秒前
栗子对应助文件撤销了驳回
11秒前
香蕉觅云的应助被李瑞程采纳,获得10
11秒前
11秒前
SciGPT的应助被有缘采纳,获得10
11秒前
爽飞边芷了完成签到,获得积分10
12秒前
香蕉寒珊的应助被王大橙子采纳,获得10
12秒前
Valerie发布了新的文献求助10
12秒前
bkagyin的应助被lmy采纳,获得10
12秒前
CodeCraft的应助被hml123采纳,获得10
13秒前
麦粑块发布了新的文献求助10
13秒前
13秒前
王萍有发布了新的文献求助10
13秒前
14秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
Improving the wear resistance of vehicle gear oil by oil soluble LaF3 nanoparticles and typical additives of gear oil 500
Boundaries and Bodhisattvas 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8183263
求助须知:如何正确求助?哪些是违规求助? 9572532
关于积分的说明 21826339
捐赠科研通 7780187
什么是DOI,文献DOI怎么找? 3433993
关于科研通互助平台的介绍 2558131
邀请新用户注册赠送积分活动 2480239