情报杂志2025，Vol.44Issue(10)：155-162,154,9.DOI:10.3969/j.issn.1002-1965.2025.10.019
基于LLM与可视化的图数据库专利知识图谱构建及分析
Construction and Analysis of Patent Knowledge Graph in Graph Database Field Based on LLM and Visualization
摘要
Abstract
[Research purpose]Addressing the issue of low efficiency in traditional patent knowledge graph(PKG)construction methods,this study proposes a patent knowledge graph construction method based on large language models(LLM),aiming to achieve intelligent processing of patent information in the field of graph databases,and thereby enabling knowledge graph-based patent analysis.[Research method]Using graph database patents as samples,this study designs LLM prompt word extraction knowledge to construct PKG.With the help of Neo4j visualization,it sorts out the technological evolution path from multiple dimensions,and analyzes the influence of applica-tion field keywords and nodes using TextRank and PageRank algorithms.[Research result/conclusion]The research shows that LLM can achieve automated construction of knowledge graphs;graph database technology has undergone three stages of development:infrastructure optimization,distributed expansion,and AI integration;the application fields are centered around knowledge graphs,data processing,and models,with knowledge graph nodes having the strongest influence.关键词
大语言模型/图数据库/专利信息/知识图谱/知识抽取/可视化Key words
large language model/graph database/patent information/knowledge graph/knowledge extraction/visualization分类
信息技术与安全科学引用本文复制引用
周贞云,黄昭昭,邱均平..基于LLM与可视化的图数据库专利知识图谱构建及分析[J].情报杂志,2025,44(10):155-162,154,9.基金项目
国家社会科学基金重大项目"基于大数据的科教评价信息云平台构建和智能服务研究"(编号:19ZDA348) （编号:19ZDA348）
浙江省科学技术厅技术开发项目"基于图技术的科教评价信息资源开发与管理"(编号:KYH263123010M)研究成果. （编号:KYH263123010M）
