大语言模型辅助的知识图谱渐进式错误修复方法  

A Progressive Error Repair Method for Knowledge Graphs Assisted by Large Language Models

作　　者：郑旭 刘静[1,2] 张栗粽 闫科 宋发仁[3] 常清雪 ZHENG Xu;LIU Jing;ZHANG Lizong;YAN Ke;SONG Faren;CHANG Qingxue(University of Electronic Science and Technology of China,Chengdu 611731,China;Kash Institute of Electronics and Information Industry,Kashi 844099,China;University of Electronic Science and Technology of China,Shenzhen Institute for Advanced Study,Shenzhen 518000,China;Sichuan Huakun Zhenyu Intelligent Technology Co.,Ltd,Chengdu 610000,China)

机构地区：[1]电子科技大学,成都611731 [2]喀什地区电子信息产业技术研究院,喀什844099 [3]电子科技大学(深圳)高等研究院,深圳518000 [4]四川华鲲振宇智能科技有限责任公司,成都610000

出　　处：《宇航计测技术》2025年第2期63-71,90,共10页Journal of Astronautic Metrology and Measurement

摘　　要：知识图谱是有效整合和组织信息的重要知识表示形式,广泛应用于搜索引擎、智能问答和推荐系统。传统知识图谱构建依赖于人工标注和规则系统,规模巨大,质量参差,难以适应海量数据的动态变化。近年来,大模型在知识生成方面表现突出,但提升知识图谱错误检测以及修正的研究仍然缺乏。为此,提出了一种大语言模型辅助的知识图谱渐进式错误修复方法。该方法利用嵌入模型评估知识三元组质量,以高质量三元组作为提示学习内容,实现了基于大语言模型的知识修复。基于大量试验分析,所提出的方法能够显著提升知识图谱的推理能力。Knowledge graph is an important form of knowledge representation, which can integrate and organize information effectively.It has been widely used in search engines, intelligent question answering and recommendation systems.Traditional knowledge graph construction relies on manual annotation and rule-based systems, which is huge in scale and uneven in quality, and is difficult to adapt to the dynamic changes of massive data.Recently, large models have shown superior performance in knowledge generation.However, there is still a lack of research on large language models to enhance knowledge graph error repairing.Therefore, a progressive error correction method for knowledge graphs, assisted by large language models, has been proposed.Using embedding models to evaluate the quality of knowledge triples and high-quality triples as prompts for learning content, knowledge correction by large language models is realized.Based on extensive experiments, the proposed method significantly enhances the reasoning ability of knowledge graphs.

关 键 词：知识图谱 大语言模型 嵌入模型 渐进式方法 错误修复 

分 类 号：TP14[自动化与计算机技术—控制理论与控制工程] V19[自动化与计算机技术—控制科学与工程]