基于工程领域知识背景的多轮对话系统  

Multi-Turn Dialogue System Based on Engineering Knowledge Background

作　　者：高岩 李晓琳 王新柏 Gao Yan;Li Xiaolin;Wang Xinbai(Heilongjiang Transportation Investment Engineering Construction Co.,Ltd.,Harbin,China)

机构地区：[1]黑龙江省交投工程建设有限公司,黑龙江哈尔滨

出　　处：《科学技术创新》2025年第4期197-200,共4页Scientific and Technological Innovation

摘　　要：本文提出一种基于工程领域知识背景的多轮对话系统,致力于将对话系统与知识检索应用于工程领域。工程领域的知识架构复杂、内容丰富,但施工范围庞大、人员众多、时常会出现无效沟通。本文通过知识图谱提取技术和句子向量化技术构建工程领域的知识库,通过知识库实现对提问的有效检索,并得到相关性较高的答案,然后基于LangChain框架和ChatGLM模型实现多轮对话问答,为多轮对话技术应用于工程领域提供了可行性方案,为促进工程领域的数字化转型做出了贡献。This paper proposes a multi-round dialogue system based on the knowledge background of the engineering field,and is committed to applying the dialogue system and knowledge retrieval to the engineering field.The knowledge structure in the field of engineering is complex,rich and requires a lot of language communication in the day-to-day operation of the enterprise.However,the scope of construction is huge,there are many personnel,and there is often ineffective communication.This paper constructs a knowledge base in the engineering field through knowledge graph extraction technology and sentence vectorization technology,realizes the effective retrieval of questions through the knowledge base,and obtains answers with high relevance,and then realizes multi-round dialogue Q&A based on the LangChain framework and ChatGLM model,which provides a feasible solution for the application of multi-round dialogue technology in the engineering field and contributes to promoting the digital transformation of the engineering field.

关 键 词：人工智能 自然语言处理 多轮对话 工程领域 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]