Research On Semantic-based Dynamic Services Binding

Currently, the problem of static service binding still exists in both traditional Web Services and Semantic Web Services, that means generally the incidence relation between service provider and service user does not change while developing and deploying applications.But actually in real service application environment, the services needed by service requirements always vary dynamically. On one hand, the old services maybe cannot be accessed by certain reason while the new services for the same function appear continuously; On the other hand, quality of service maybe change with network or service load dynamically.This paper is mainly focus on how to help service user bind best-quality service dynamically in the changing environment. It gets the following achievements:1. Dynamic Services Binding (DSB for short) algorithm is proposed based on web-service quality evaluation model and load-sharing strategy.DSB considers the services with higher service quality range first, and chooses the service whose load is not overweight as the best binding service using Round-Robin algorithm.The experiment results show the efficiency and correctness of DSB algorithm.2. One improved Web Services architecture is proposed: add service proxy agent between service provider and service user. Service users send semantic request description to the service proxy and then the proxy will follow up to look for services, match them and use DSB to bind them. After the binding, service proxy will notify the service user in time. And the user will send the service calling parameter to the proxy and the proxy will really call target service and offer the feedback to service user.The improved Web Services architecture can resolve the problem of"static service binding"efficiently.3. According to the improved service architecture, one prototype system is designed and implemented based on the DSB model. The prototype system mainly includes five function models: use protege tool to set up domain ontology and OWL-S Editor service ontology; implement the interface of semantics service register center; generate description from service request; find and match service; bind service dynamically.