论文部分内容阅读
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。
The UMLS semantic proposition is the smallest unit of semantic knowledge represented by a triad. The subject and object are all concepts in the UMLS super-thesaurus, and the predicate is the semantic relationship in the UMLS semantic network. The extraction process of UMLS semantic proposition involves the aspects of shallow syntactic analysis, concept mapping, predicate recognition and semantic proposition generation. Two medical information resource aggregation methods based on UMLS semantic proposition --- the aggregation method using knowledge units as resource units and the document association data as resource unit aggregation methods, the aggregation results are knowledge network and document network respectively.