The Semantic Service Search Engine (S3E)

Currently, the Web is an important part of people's personal, professional and social life and thousands of services are becoming available online to support these. Since 2005, many efforts have been made to semantically describe Web services and several models have been proposed towards this direction, e.g. OWL-S, WSMO, SAWSDL. The Web follows a decentralized architecture, thus all the services are available at some location; but finding this location remains an open issue. Many efforts have been proposed to solve the service search problem. In this work, an innovative approach for semantic service search is proposed. Our effort addresses mainly two shortcomings. First, it does not require from the service providers to publish their services or their descriptions in a centralized service registry. Second it exploits the semantic information that exists in semantic service descriptions in order to improve service search. Our approach comprises of three main phases. The crawling phase, during which semantic service descriptions that are online are retrieved and stored locally. The homogenization phase when the semantics of every description are mapped to a Reference Service model and. Finally, the search phase when the users are enabled to query the underlying repository and find online services.