A HYBRID FUZZY MULTI-CRITERIA DECISION MODEL FOR CLOUD SERVICE SELECTION AND IMPORTANCE DEGREE OF COMPONENT SERVICES IN SERVICE COMPOSITIONS

摘要

With the recent paradigm shift towards Cloud computing and Service Oriented Architecture (SOA), Service selection and evolution have emerged as significant challenges for service integrators and maintainers. Service selection process involves both subjective and objective factors based on user feedback and performance assessment, along with inherently imprecise data. Similarly, determining the relative importance of component services (CS) in a service composition is not a trivial task, and encounters similar challenges as for service selection. This relative importance of CSs is crucial to quantify the impact of a proposed change in service composition. However, existing literature lacks a systematic procedure which aggregates user feedback and real world performance assessment data, while incorporating inherent fuzziness of these criteria. In this study, we formulate these problems as Multi-Criteria Decision problems, and propose a hybrid MCDM model based on Fuzzy Delphi methodology (FDM), Fuzzy Analytic Hierarchy Process (FAHP), Fuzzy Vikor and Fuzzy TOPSIS to tackle previously mentioned issues. The proposed model is different from existing studies in services computing, as it incorporates both user feedback and real world performance assessment data, and deals with uncertainty in decision making process. FDM is used to determine critical subjective and objective factors for service selection and importance degree of CSs in existing compositions. Then, Fuzzy AHP is used to determine the criteria weights, while FTOPSIS or FVTKOR is used to rank the alternatives.