Web Service Selection Based on Analytical Network Process Approach

Abstract

The Web service selection is a complex decision process. It engages multiple criteria for the selection of candidate service from a set of available services. There may be more than one web service meeting the functional requirements. The nonfunctional properties of these matching Web services may have varying values, which should be necessarily above minimum requirements. The selection criteria may have an interdependent relationship between them. The complexity involved in the selection process necessitates the use of quantitative techniques for decision-making. Analytical Network Process (ANP) is a quantitative approach to deal with interdependent relationships within a multi-criteria decision-making framework. This paper explains the use of an ANP for Web service selection. We propose a network model with clusters of elements relevant for Web services and solve it as an illustrative case for ANP based Web service selection.