Welcome to Service Detective
The Web is moving from a collection of static documents to a collection of services. Web Services are today accepted as a common standard for realizing service interchanges in a business to business context. Furthermore they are widely seen as the most promising technology to enable interoperability within service-oriented architectures (SOA). Current SOA solutions are mainly restricted to intra- organizational environments, and are not yet widely deployed in external settings.
The discovery of Web Services is essential for creating and utilizing dynamically created applications. Current technologies provide means to describe services on a syntactical level, thus providing only limited automation support. For efficient service discovery, a good characterization of functional and non-functional properties is though essential.
The FIT-IT project RW² (Reasoning With Web Services) has emphasized the role of semantics in service discovery and has focused on how to build and reason over formal semantic service descriptions. Semantic descriptions increase the accuracy of service discovery, however the availability of semantic service description respectively their creation constitutes a severe bottleneck. The availability of trained experts and the costs involved do not allow current approaches to scale up with the existing number of services in the Web. The Service-Detective project will take existing semantic description methodologies and focus on how they can be combined with Web principles in order to scale in a Web context.
The aim of the Service-Detective project is to make Web Services available to potential consumers in a similar way as static Web documents are made available to users by current state-of-the-art search engines. The project will deliver a search engine that will crawl the Web for services and that will thus not be limited to explicitly published services. The crawler will identify Web Service descriptions and gather related information about the services. All available information will be analyzed and added to a cohesive semantic model. On top of this semantic model reasoners can perform inferences in order to provide meaningful ranking and clustering of information. In particular Service-Detective will focus on integrating dynamic aspects of the service requesters and on building meaningful clusters of services.
Service-Detective aims at exploiting both the structure and the semantics of Web Service descriptions in order to find a balance between complex semantic query languages and keyword based approaches.