Research Of Agent-Based And QoS-Driven Web Service Discovery

Web service is a hot research at present. As one of the important branch in web service system, the development of web service discovery affects the development of the web service. There are still some weaknesses in today's web service discovery technology: In one hand, the model of web service discovery is not very perfect, it does not considering QoS authentication and user's feedback information at the some time, meanwhile, it does not considering the provider's intention while updates the web services'QoS. In another hand, nowadays'web service discovery methods based on QoS does not consider the fuzzy QoS attributes and does not support fuzzy QoS attributes'assessment and discovery.After analysis the existing model, this dissertation based on the classical web service model of SOA, extens the UDDI register center in order to make it supports the QoS, adds a QoS authentication and feedback module in the model and put forward an agent-based and QoS-driven framework. In order to dear with the problem that nowadays'algorithms do not support the fuzzy QoS attributes, we use interval numbers to describe the fuzzy QoS attributes and put forward a new algorithm which supports the fuzzy QoS attributes'assessment and discovery. We exploit a prototype system to validate the feasibility of the framework. Meanwhile, we have done some simulation experiments, the results show that the algorithm we put forward can increase the precision and the recall-precision of web service discovery.