WSTRank: Ranking Tags to Facilitate Web Service Mining

Web service tags, terms annotated by users to describe the functionality or other aspects of Web services, are being treated as collective user knowledge for Web service mining. However, the tags associated with a Web service generally are listed in a random order or chronological order without considering the relevance information, which limits the effectiveness of tagging data. In this paper, we propose a novel tag ranking approach to automatically rank tags according to their relevance to the target Web service. In particular, service-tag network information is utilized to compute the relevance scores of tags by employing HITS model. Furthermore, we apply tag ranking approach in Web service clustering. Comprehensive experiments based on 15,968 real Web services demonstrate the effectiveness of the proposed tag ranking approach.