A Proactive Approach to Semantically Oriented Service Discovery

Abstract

Web Service based computing has evolved immensely in recent years, supported by standards bodies, academic research and industry alike. One of the noticeable omissions from the web services architecture is that of standards to support automatic discovery, automatic composition and invocation of web services. While semantic web service discovery implementations exist research into autonomous semantically oriented service discovery is far from an efficient, complete solution to the problem. Current web-service capability matching research and implementation focuses on semantically enhancing the UDDI standard for web-service discovery. While this is definitely a step in the right direction, dependence on the UDDI standard may restrain capability matchers as they mature. A prime example of this is the request/response nature of a UDDI look-up. This dissertation researches an alternative approach to web service discovery that proactively informs interested participants of the availability of new services that match expressed capability requirements. By evaluating and semantically enhancing a wide area notification system this dissertation develops a scalable publish/subscribe platform for OWL-S service discovery that incorporates efficient content-based routing and an expressive subscription language. Our evaluation shows that many of the documented optimisations for text-centric content based routing actually hold in routing for more complex OWL-based concepts. We conclude that a publish/subscribe model for semantic service discovery is feasible and can potentially provide a pro-active discovery environment for human and autonomous agents alike. However, in order to realise this vision, much research in the area of distributed knowledge base consistency, ontology alignment and intelligent sharing of OWL information is needed.