An Automated Approach for Web Services Architectural Style Selection

Abstract

Selection of an appropriate architectural style is vital to the success a web service. The nature of architecture design and selection for service oriented computing applications is quite complex as compared to traditional software architecture. Web Services have complex and rigorous architectural styles with their own underlying architectural characteristics. Due to this, selection for accurate architectural style for web services development has become more complex decision to be made by architects. Architectural style selection is a multi-criteria decision and demands lots of experience in service oriented computing. There is a huge gap for automated selection of web services architectural styles. Decision support systems are good solution to simplify the selection process of a particular architectural style. Our research suggests an automated approach using DSS for selection of architectural styles while developing a web service to cater FRs & NFRs (Functional & Non Functional Requirements). Our proposed mechanism helps architects to select right web service architectural pattern according to domain, and non-functional requirements without compromising quality. In this paper a rule base DSS has been developed using CLIPS (C Language Integrated Production System) to support decision process in multi-criteria requirements. To select suitable web service, system takes architectural characteristics, domain requirements and software architect preferences for NFRs as input by applying rule base approach. Next Weighted Sum Model has been applied to prioritize quality attributes and domain requirements. Scores are calculated using multiple criterions to choose the final architecture style.