Combining skyline and multi-criteria decision methods to enhance Web services selection

Abstract

This paper describes a new approach based on combining three Multi-Criteria Decision Making methods: Skyline, AHP and Promethee, for ranking Web services. The Skyline is used to reduce the search space and focusing only on interesting web services that are not dominated by any other service. AHP is used to weight QoS criteria in a simple and an intuitive manner; it allows for checking whether the weights assigned are consistent. Promethee method is leveraged to rank skyline services, by taking advantage of the outranking relationships between skyline candidate services and generating positive, negative and Net flows. An algorithm is proposed to rank skyline services based on Net flow. A case study based on an example of QoS dataset is presented to illustrate the different steps of our approach. The experimental evaluation conducted on real Dataset demonstrates that our approach can better capture the user preferences and retrieve the best ranked Skyline services.