Putting Together Skyline and Agent Paradigms to Enhancing Web Service Composition

The fast growing number of alternative web services that provide the same functionality but differ in quality parameters and their continuous updating pose a new challenge for service composition: selecting proper services that achieve a composition with the best user desired preferences. However, the total number of possible compositions is often extremely high, an online exhaustive evaluation of all solutions involves a poor performances. In this paper, we address this issue and propose a solution for web service selection by leveraging the skyline operator and the agent paradigm. The solution allows providing a composite service which better satisfies user´s preferences. User preferences are modeled using fuzzy sets. First, relevant services are obtained by applying the skyline operator. Then we select the top-k services by ranking skyline services using a fuzzy dominance method. Thereafter, we sort all the combinations of composite solutions according to their global satisfaction, and finally the best composite service is retained. In our approach, the composite service can be described in a sequential or/and a parallel structure. A set of experiments are done to show the effectiveness and the performance of the approach on a large number of web services.