A Specialized Search Engine for Web Service Discovery

Web service discovery on the web is not a trivial task as the number of available web service descriptions continuously increases, and global UDDI registries are no longer available. As discovery through conventional, general-purpose search engines does not yield satisfactory results, a more promising alternative should be explored through specialized search engines. This paper explores the design and implementation of such a framework, resulting in WESS, a search engine targeted to discovering and retrieving web service descriptions. The presented system features an adaptive web service description collection process, through specialized and directed crawling, as well as an enhanced indexing and retrieval mechanism, which handles description documents as semi-structured text, separating actual information from tags and annotations. The paper also presents experiments and use cases regarding different search scenarios, in addition to performance results.