摘要

Web services composition has received much interest for its ability to effectively integrate distributed and heterogeneous applications. With the popularity of Web service, a wide variety of Web services with similar functions are offered and there can be more than one combination of Web services for a complex task, therefore, there is great need to distinguish them by using a set of well-defined quality of service. To accurately define the quality of a composite Web service, an extended quality model for the composite Web service is proposed, which includes the general quality criteria, message compositionality and composition rationality between Web services. Based on the extended quality model, a multi-objective genetic algorithm for optimizing the Web services selection is proposed, which can handle global user's requirements and constraints. The experimental results show the multi-objective genetic algorithm proposed can effectively optimize the services selection for the composite Web service. Moreover, the experimental results also indicate that the approach has better scalability and can be applied to a large-scale distributed service computing environment.