Skip to Main Content
Due to userÂ¿s inability to define his service requirement in a precise way, the services returned by service search engine (SSE) are mostly inadequate or inaccurate. Taking a clustering method on its potential search results would not only improve its search utility , but also enhance its effectiveness and accuracy than traditional ranked-list style in helping users to find relevant services . For a clustering mechanism to be effective, a retrieval model on which it applies must also be well defined. In this paper, we will first introduce our prototype semantic service retrieval engine, and then propose an efficient clustering algorithm for organizing returned services. We evaluate the proposed method with a well-known service collection: OWL-S TC. Thorough experiment show positive results, which demonstrate the feasibility of our approach.
Date of Conference: 3-5 Dec. 2008