基于GraphRAG的物流知识问答系统应用研究

打开文本图片集

中图分类号：TP391；TP182 文献标识码：A 文章编号：2096-4706（2025）20-0040-04

Abstract： In view of the problems existing in traditional logistics knowledge processing methods，such as knowledge extractionrelying on manual labeling，high cost and time consumption，this paper proposes a logistics Knowledge Graph construction method based on a Large Language Model.By designing prompt words and with the help of the semantic understandingandreasonngcapabilitiesoftheLargeLanguageodel，knowledgeisextractedfromloisticselateddatadthe extracted knowledge isstored intheNeo4j graphdatabase.Inthe implementationofthequestionansweringsystem，GraphRAG technologyisusedtoretrieverelevant entities，elationshipsandtheiratributesfromthelogistics KnowledgeGraph，soas to enhanceandoptimize the prompt words，providereliablebackground knowledge support fortheLarge Language Model to generateanswers，and efectivelyreducetheilusion phenomenonofLarge Language Model.System testsshowthatthismethod effctivelyalleviates the problems ofknowledge fragmentationandlogicaljumpin traditional questionansweringsystems.

KeyWords： logistics knowledge; question answering system; Large Language Model; Knowledge Graph; GraphRAG

0 引言

随着行业数字化转型加速，知识问答系统成为提升效率的关键工具。（剩余5211字）