基于本体的概念相似度计算及其应用  

Concept similarity computation based on ontology and its application

作　　者：冉婕[1,2] 孙瑜[1] 漆丽娟[2] 

机构地区：[1]云南师范大学计算机科学与信息技术学院,云南昆明650092 [2]云南昭通师范高等专科学校计算机科学系,云南昭通657000

出　　处：《微型机与应用》2010年第11期14-16,20,共4页Microcomputer & Its Applications

基　　金：国家自然科学基金项目(60903131);云南省社会发展科技计划应用基础研究项目(2009ZC052M);云南省教育厅重点项目(07Z10661)

摘　　要：提出了基于语义相似度和相关度的综合概念相似度计算方法。语义相似度考虑了语义距离和本体库特征,加入概念的信息量、概念的深度、概念的密度和不对称因子的辅助影响;语义相关度从直接相关、间接相关、直接继承和间接继承几个方面考虑。通过实验和两种传统的语义相似度计算方法进行对比,本方法能更好地区分本体树中不同关系的概念对,验证了该方法的有效性。This paper puts forward an integrated method based on semantic similarity and semantic relevancy. In semantic similarity, we think about the semantic distance and the characteristics of ontology, the concept＇s amount of information, the depth of concept, the density of concept and symmetry factor; in semantic relevancy, we think about the directly relation and indirectly relation, direct inheritance and indirect inheritance. Compated with two traditonal semantic similarity computation methods, this method can better distinguish the different concept in ontology tree and this method is effective.

关 键 词：本体 相似度 语义距离 信息量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]