文章摘要
虞红蕾,曹灵勇,瞿溢谦,等.消渴病经方知识图谱构建与知识发现[J].浙江中医药大学学报,2022,46(2):113-119.
消渴病经方知识图谱构建与知识发现
Knowledge Graph Construction and Knowledge Discovery of Classic Prescriptions of Diabetes Disease
DOI:10.16466/j.issn1005-5509.2022.02.001
中文关键词: 经方  消渴病  人工智能  知识图谱  知识发现  中医药
英文关键词: classical prescription  diabetes disease  artificial intelligence  knowledge graph  knowledge discovery  traditional Chinese medicine
基金项目:浙江中医药大学横向(涉企)项目(2020-HT-161、2020-HT-837)
作者单位
虞红蕾 浙江中医药大学基础医学院 杭州 310053 
曹灵勇 浙江中医药大学基础医学院 杭州 310053 
瞿溢谦 浙江中医药大学基础医学院 杭州 310053 
刘畅 广东省中医院 
杨帆 杭州甘之草科技有限公司 
王平 杭州甘之草科技有限公司 
王磊 浙江中医药大学基础医学院 杭州 310053 
林树元 浙江中医药大学基础医学院 杭州 310053 
摘要点击次数: 2123
全文下载次数: 1010
中文摘要:
      [目的] 构建经方领域的消渴病知识图谱,并在此基础上进行知识查询和发现。[方法] 筛选经方古籍中消渴病的相关内容,通过本体七步法构建消渴病本体,以结构化的三元组数据来表示研究内容,并进行语义消歧,用Excel软件存储结构化数据,再导入Neo4j图数据库,构建消渴病的经方知识图谱。[结果] 构建了包含1 432个节点、3 067个关系的消渴病经方知识图谱,其模式层包含24个节点、24个标签、54条关系和24种关系类型,通过Cypher语言可以从辨证、立法、处方、遣药四个方向进行描述性检索与知识发现。[结论] 以知识图谱的方式结构化展示消渴病经方相关内容,模拟经方理论体系与临床诊疗思维,通过实现复杂知识之间的多途径联系,有助于更深入地进行知识发现与探索。
英文摘要:
      [Objective] To construct the knowledge graph of diabetes disease in the field of classical prescriptions, and to query and discover the knowledge on this basis. [Methods] The content of classic prescriptions related to diabetes disease was screened, the ontology of diabetes disease was constructed by domain ontology seven-step method, the research content was represented by structured triad data and semantic disambiguation was carried out, the structured data were stored by Excel software, and then imported into Neo4j graph database to complete the knowledge graph of classical prescriptions of diabetes disease. [Results] The knowledge graph of classic prescriptions of diabetes disease was constructed with 1 432 nodes and 3 067 relationships. Its schema layer contains 24 nodes, 24 labels, 54 relationships and 24 relationship types, and the descriptive retrieval and knowledge discovery were carried out from the four directions of syndrome differentiation, principles establishment, prescription and herbs usage through Cypher language. [Conclusion] To display the relevant contents of diabetes disease classic prescriptions structurally in the way of knowledge graph, simulate the theoretical system of classic prescriptions and clinical diagnosis and treatment thinking, more in-depth knowledge discovery and exploration can be carried out by realizing the multi-channel connection between complex knowledge.
查看全文   查看/发表评论  下载PDF阅读器
关闭