Web Service Selection Approach Based on Agent and Fuzzy Logic-

The context of this work is the selection of the services among a list of similar services. To differentiate between similar web services, the best approach to use is the one based on service quality (QoS) criteria. However, these criteria most of the time are imprecise and sometimes uncertain and ambiguous because they are based on discrete numerical data. In some situation, we need to apply the concept of partial truth, where the partial truth values are between two numerical values. The fuzzy logic can be applied to support the representation of imprecise, uncertain and ambiguous representation. It is based on the observation that people make decisions based on imprecise and nonnumerical information. In this work, we propose an approach for selecting web services based on the fuzzy logic. Indeed, we use fuzzy logic by translating the QoS values that are attributed to services into linguistic terms. In addition, we also allow the service client to formulate its requests in the form of linguistic terms. In this process, the system selects and render the desired web service to the client. Our solution is represented as three layers, i) User Layer, that receives functional or fuzzy requests from clients which treats and forward them to the next layer, ii) Basic Layer that implements the service-oriented architecture and also, obtains the QoS values from service repository; and iii) Fuzzy Layer that takes as input the quality criteria of each web service then proceeds to their fuzzification, ranking and finally assign the scores to the different web services. To validate our solution, we implemented it as a multi-agent system.