Modeling Feature Interactions among Web Services based on Feature Models

Abstract

The online community has been growing steadily year on year and with this escalation the number of services provided has increased in an attempt to meet the demands of a computer literate audience. This has led to a progression from a human-oriented use of the web to an application-oriented structure through the use of web services composed in various ways. The web service composition process potentially involves a large number of interactions among the features of the services involved in the process. This demands a flexible modeling approach that should be capable of modeling and analyzing the interactions among such features. The specified model should be verified to guarantee that the designed service composition process yields feature interaction free services. However none of the traditional approaches (BPEL, WSCDL, OWL-S, and WSMO etc.) offers any direct support for verification of service composition at design time to evaluate its correctness. In this paper we conducted a substantial survey of current approaches used for the analysis of feature interaction in web services and an integrated approach is proposed to handle this issue in a more flexible way, by using feature modelling and model checking techniques. We suggest that the proposed approach is more scalable as compared to the traditional approaches surveyed.