一种基于概念格属性约简的语义相似度计算方法  

A Method of Semantic Similarity Computing Based on Attributes Reduction of Concept Lattice

作　　者：伍振兴[1] 

机构地区：[1]琼州学院电子信息工程学院,海南三亚572022

出　　处：《琼州学院学报》2012年第2期43-46,共4页Journal of Qiongzhou University

摘　　要：提出一种基于概念格属性约简的语义相似度综合计算方法。该方法要考虑概念在概念格中的语义距离、内容以及层次结构三个方面的因素对语义相似度的影响。同时对所构造的概念格进行属性约简,删除本体合并中的冗余信息,约简绝对不必要属性,以最少的信息精确地反映概念格及本体的本质和结构。最后通过实例分析此方法的可行性。This paper presents an semantic similarity computing method based on concept lattice＇ s attribute reduction, The method is to consider the concepts in the concept lattice of semantic distance, content and hierarchy structure of three factors on the effect of semantic similarity ; With this background we construct the concept lattice, then we conduct attribute reduction on the concept lattice, remove redundant information of the ontology merging, reduce the unnecessary attribute, thus to use the minimum information to accurately reflects the nature and structure of the concept lattice and the body. Finally, we use example analysis to prove the feasibility of this method.

关 键 词：概念格 属性约简 语义相似 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]