Research On Optimization Of Context-aware Web Services Composition

With the rapid development of Internet technology, Web service technology has been widely applied. Since the single Web service with single function can not meet users’ various requirements, the concept of Web services composition has been proposed. Several problems exist in recent Web services composition: Firstly, because there are a lot of services in the Internet, which have similar or even the same functions, the process of searching single service and selecting service composition is time-consuming and the efficiency of services composition is not high. Secondly, in the Web services composition, we often only consider the service functional requirements and a part of non-functional requirements and not consider the non-functional information of services, the correlative information of services and the user preferences information, so the quality of services composition is not high. In addition, in the process of formulating services composition plan, we often do not consider the correlative information among services, so in the actual operation, the success rate of the services composition plan is not high. In view of the above problems, in this paper, the context-aware technology is introduced into the process of Web services composition plan selection. This paper considers all the context information of Web services and their compositions in the process of Web services composition, in order to optimize the efficiency, quality and success rate of Web services composition.This paper presents an optimizational method of context-aware Web service composition. This paper defines the concept of service correlation context, and divides it into statistics correlation context, quality correlation context and interface correlation context. In the process of services composition plan selection, the system chooses the best scheme through the analysis of service context, service correlation context and users’ context information. Firstly, statistics correlation context is applied to the generation process of candidate services composition plan set, and as a result, the efficiency of searching candidate services composition plan is improved. Secondly, quality correlation context and interface correlation context are applied to the selection process of the final services composition plan, and as a result, the quality and success rate of services composition plan is improved. This paper use genetic algorithm to solve the optimization of context-aware Web services composition. Considering the service correlation context, this paper improves the population initialization, coding method, fitness function and genetic operation of genetic algorithm. Finally, experiments validate that the optimizational method for context-aware Web service composition given in this paper is effective.