Ranking semantic web services using semantic-distance information

Abstract

World wide web will gain its full power when semantic meta-data descriptions are added and automatic processing of the web is realized. Web services gained importance in the e-commerce era in recent years. As the number of web services increase, discovering correct web services for user needs becomes a problem. An effective discovery is only possible when there is semantic information. Semantic web enables automatic discovery, invocation, composition and mediation of web services by users, agents and programs. The first step in semantic web is the effective discovery of web services. Most of the current solutions and approaches on web service discovery and selection are limited in the sense that they are strictly defined, and they do not use the full power of semantic and ontological representation. This research proposes a general “semantic web service discovery framework” in which: Users or software agents should able to set the selection criteria as XML documents, defined in preliminary format, in an extensible and modifiable manner. A Matchmaking Agent (MS-Matchmaker), which is the main component of the framework, processes the service discovery query. The matchmaking is done via an efficient matching algorithm. The matchmaker will return a rated and ordered set of suitable web services to the service requestor who may be a human-user, software agent, component or other. Semantic distance of ontology concepts are the main criteria of rating the web services in parallel with user definitions, which is different from other approaches.