以UMLS语义命题为基础的医学信息资源聚合  

Polymerization of Medical Information Resources Based on UMLS Semantic Propositions

作　　者：郭少友[1] 李庆赛[1] 

机构地区：[1]郑州大学信息管理学院

出　　处：《图书情报工作》2014年第3期99-105,共7页Library and Information Service

基　　金：教育部人文社会科学研究规划基金项目"图书馆数字资源的细粒度语义化描述与复用研究"(项目编号:13YJA870008)研究成果之一

摘　　要：UMLS语义命题是用三元组表示的最小语义化知识单位,其主语和宾语都是UMLS超级叙词表中的概念,谓词是UMLS语义网络中的语义关系。UMLS语义命题的抽取过程涉及浅层句法分析、概念映射、谓词识别与语义命题生成等环节。两种以UMLS语义命题为基础的医学信息资源聚合方法———用知识单元作为资源单位的聚合方法和用文档关联数据作为资源单位的聚合方法,其聚合结果分别是知识网络和文档网络。UMLS semantic propositions are the smallest units of semantic knowledge represented as triples, in which subjects and objects are concepts from the UMLS metathesaurus and predicates are semantic relations from the UMLS se- mantic network. The extraction process of UMLS semantic propositions involves the shallow parsing, concept mapping, predicate identification and generation of semantic proposition. This paper proposes two polymerization methods of medical information resources based on UMLS semantic propositions, which respectively regard knowledge units and linked docu- ment data as resource units. The polymerization results are respectively a knowledge network and a document network.

关 键 词：UMLS 语义命题 医学信息资源 聚合 

分 类 号：G254[文化科学—图书馆学]