An Approach to Web Services Selection for Multiple Users

Web service selection for multiple users is an important aspect for achieving efficient operations for web service applications. Its aim is to select optimal solutions, in which each abstract web service in the workflow of a web application is bound to its corresponding concrete web service with the optimal quality of service (QoS), for all users based on their QoS requirements for the workflow. There are a lot of approaches to resolve this problem, but they do not consider each user's different QoS requirements or have prohibitively large overhead for using these approaches. In this paper, we present an approach to significantly improve the efficiency of web service selection by the advanced a-fully polynomial time approximation scheme to calculate the Pareto optimal set, where each solution is not dominated by others. Additionally, this approach reduces its overhead further by adopting artificial bee colony algorithm to select an optimal solution from the Pareto set for each user. Experimental results are presented to show the efficiency of this approach.