Abstract:

Objective To construct an intelligent question-answering system for breast cancer health information based on knowledge graph，aiming to improve the self-management awareness and ability of breast cancer patients. Methods The team systematically retrieved breast cancer-related guidelines and expert consensuses published between January 2019 and December 2023，and referred to professional books and other materials to build a knowledge graph of breast cancer health information through knowledge extraction，knowledge fusion，and knowledge storage process. From April to December 2024，we established a breast cancer health information knowledge base，realized the matching of user questions and knowledge graph content through entity link model，completed the development of knowledge query and display functions，and finally formed an intelligent question-answering system. From January to February 2025，13 breast cancer patients and 4 clinical medical staff were selected as research subjects to carry out the preliminary application and usability evaluation of the system. Results An intelligent question-answering system for breast cancer health information with the ontology layer covering prevention，screening，symptoms，examination，treatment，complications，reexamination，and health management was successfully constructed. The usability evaluation showed that the standardized average score of the system by patients and medical staff was（82.21±19.32） points，indicating that the system usability was at a high level. Conclusion The intelligent question-answering system for breast cancer patients’ health information based on knowledge graph can provide patients with accurate and professional health information with strong feasibility.

Key words: Breast Cancer, Health Information, Knowledge Graph, Intelligent Question-Answering System, Nursing Care

摘要：

目的 构建基于知识图谱的乳腺癌患者健康信息智能问答系统，旨在为提升乳腺癌患者的自我管理意识与能力提供借鉴。方法 系统检索2019年1月—2023年12月发布的乳腺癌相关指南、专家共识，并参考专业书籍等资料，通过知识抽取、知识融合、知识存储流程，构建乳腺癌患者健康信息知识图谱。2024年4—12月，建立乳腺癌患者健康信息知识库，通过实体链接模型实现用户问题与知识图谱内容的匹配，完成知识查询与展示功能开发，形成智能问答系统。2025年1—2月，研究团队选取13例乳腺癌患者及4名临床医护人员作为研究对象，开展系统初步应用与可用性评价。结果 构建了以预防、筛查、症状、检查、治疗、并发症、复查、健康管理为本体层的乳腺癌患者健康信息智能问答系统。可用性评价显示，乳腺癌患者与医护人员对系统可用性评价标准总分为（82.21±19.32）分，系统可用性较好。结论 基于知识图谱的乳腺癌患者健康信息智能问答系统可向患者提供准确、专业的健康信息，具有较好的可用性。

关键词: 乳腺癌, 健康信息, 知识图谱, 智能问答系统, 护理