Using semantic information for distributed web service discovery

Abstract

With the increase in number and size of service registries, web service discovery becomes a challenging activity. There is also tremendous increase in the number of web services that are not registered to any of the business registries. Under these conditions, finding an appropriate web service may lead to problems in terms of performance, efficiency, end-to-end security and quality of the discovered services. Use of semantic information has been studied in web service discovery for improving the accuracy of the service discovery task. However, semantic information can be useful for improving the time efficiency as well. In this work, we propose a web service discovery architecture that is based on a coordinated set of domain-specific service discoverers crawling both business registries and private sites. By distributed and domain-specific crawling, it is aimed to shorten the search time and to provide up-to-date service status and load balancing via scalability.