乳腺癌癌因性疲乏管理知识图谱的构建  

Construction of a knowledge graph for management of cancer-related fatigue in breast cancer survivors

作　　者：丁元旗 黄青梅 吴傅蕾 王玲 杨瑒[2] 李军 袁长蓉 Ding Yuanqi;Huang Qingmei;Wu Fulei;Wang Ling;Yang Yang;Li Jun;Yuan Changrong(School of Nursing,Fudan University,Shanghai 200032,China)

机构地区：[1]复旦大学护理学院,上海200032 [2]复旦大学附属肿瘤医院 [3]南京大学社会学院

出　　处：《护理学杂志》2024年第17期10-14,共5页Journal of Nursing Science

基　　金：国家自然科学基金资助项目(72374048);上海市卫生健康委员会资助项目(20234Y0088)。

摘　　要：目的构建乳腺癌癌因性疲乏管理的知识图谱,为健康领域个性化推荐系统和智能问答系统的开发奠定基础。方法采用自顶向下的方法构建乳腺癌癌因性疲乏管理知识图谱。首先,采用文献研究法建立疲乏管理知识库,为图谱构建提供知识来源;其次,构建知识图谱的本体模式层,完成知识建模;再次,采用人工方法基于本体模式层对知识库中的内容进行知识抽取,并以三元组标准格式将非结构化文本数据转化为结构化数据;最后,采用MongoDB数据库和VUE前端技术实现知识图谱的可视化呈现。结果共纳入11篇指南、4篇专家共识、6篇证据总结作为图谱的知识来源。图谱本体模式层包括8个实体类型和关系类型,人工抽取124个实体,形成121组三元组,最终完成乳腺癌癌因性疲乏管理知识图谱的可视化。结论乳腺癌癌因性疲乏管理知识图谱以图形形式呈现疲乏管理知识脉络,能够提供知识存储和知识搜索功能。Objective To construct a knowledge graph for management of cancer-related fatigue in breast cancer survivors,and to lay a foundation for developing personalized recommender system and intelligent question answering system in the health field.Methods A top-down approach was used to construct the knowledge graph.First,we constructed the data resource by evidence-based literature review;second,we built domain ontology to normalize the knowledge of cancer-related fatigue management;third,we ma-nually extracted knowledge from the data resource and translated the knowledge from unstructured text to structured data as"Entity-Relation-Entity"triples;and finally,we used the MongoDB database and Vue.js for knowledge visualization.Results A total of 11 guidelines,4 expert consensus statements,and 6 evidence summaries were included for data resources.The domain ontology covered 8 entity types and 8 relationship types.A total of 124 entities were manually extracted to form 121 triples.Finally,we developed the visual presentation of the cancer-related fatigue management for breast cancer survivors.Conclusion The knowledge graph for management of cancer-related fatigue in breast cancer survivors can present the relationships among fatigue management knowledge,and is useful in knowledge storage and knowledge retrieval.

关 键 词：乳腺癌 癌因性疲乏 疲乏管理 症状管理 知识图谱 人工智能 数智化 肿瘤护理 

分 类 号：R473.73[医药卫生—护理学] R737.9[医药卫生—临床医学]