一种基于本体概念语义距离的服务相似度度量方法  

Service Similarity Measure Based on Semantic Distance of Ontology Concepts

作　　者：葛继科[1] 邱玉辉[2] 

机构地区：[1]西南大学计算机与信息科学学院,重庆400715 [2]西南大学语义网格实验室,重庆400715

出　　处：《计算机科学》2009年第6期181-184,共4页Computer Science

基　　金：973国家重大基础研究项目(2003CB317008);西南大学研究生科技创新基金(2006011)资助

摘　　要：随着语义Web服务及语义网格服务应用的不断深入,对服务资源的需求日益增长,服务匹配在服务发现和服务组合研究中的地位也日渐重要。在服务使用OWL-S描述的前提下,服务匹配通常认为是本体概念的匹配,概念匹配的目的是发现概念间的语义相似度。概念的语义相似度不但与概念间的距离有关系,而且还受概念在本体中层次深度的影响。综合考虑这两个因素,提出了一种基于语义距离的概念相似度度量方法,给出了语义距离的定义,明确了语义距离与语义相似度的关系。最后,通过与其他方法的实验比较,验证了该方法的有效性。With the application of semantic Web service and on the fly, and service matching is more and more important semantic Grid service, the requests of service resource are in the domain of service discovery and service composition. Under the precondition of service described by OWL-S, service matching are generally considered as the matching of on- tology concepts, and the purpose of concepts matching is that finding semantic similarity between concepts. The seman- tic similarity between concepts is influenced not only by the distance between concepts, but also by position of concept in the ontology hierarchy architecture. In this paper, considering the above two factors, we proposed a concepts similarity measure based on the semantic distance, defined the semantic distance, and stated the relationship between the semantic distance and semantic similarity. At last, we also provided an experimental comparison of our measure against traditional similarity measures, and proved empirically the benefits of our approach.

关 键 词：服务匹配 语义相似度 本体 语义距离 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]