基于MeSH的主题结构分析方法研究  

Research on the Method of Topic Structure Analysis Based on MeSH

作　　者：安新颖[1] 

机构地区：[1]中国医学科学院医学信息研究所,北京100005

出　　处：《情报理论与实践》2009年第5期99-102,共4页Information Studies:Theory & Application

基　　金：中央公益性科研院所基本科研业务费课题"医学信息分析评价的模式探讨与方法应用研究"资助的成果之一;项目编号:07R0203

摘　　要：本文充分考虑了主题词之间的已知关联和未知关联,利用MeSH词表对已知关联进行了处理,优化了主题结构分析方法,并以属分关系为例,对该方法进行了实证分析。结果表明在阈值一定的前提下,基于MeSH的主题结构分析方法能够有效地剔除词间的已知关联,揭示出相对较微弱的词间未知关联,起到了主题词关系过滤的作用,为知识发现奠定了基础。Taking full account of the known and unknown associations between subject headings, this paper processes the known associations with Medical Subject Heading （MESH） and optimizes the method of topic structure analysis. The paper gives an empirical analysis of the method based on hierarchical relationship. The result shows the method of topic structure analysis based on MeSH can pick out the known associations between subject headings effectively in a certain threshold, and disclose the unknown weak associations between subject headings. The method plays a role in filtering relationships between subject headings and lays the foundation for knowledge discovery.

关 键 词：医学主题词 主题结构 分析方法 弱关联 

分 类 号：G353.1[文化科学—情报学]