A Multi-agent-based Qos-driven Web Service Discovery and Composition Framework

Web service has been playing a magnificent role in the field of application development. It facilitates giving a global touch to standalone applications components to interact with each other via interfaces and form larger application systems. Web service users who seek for single web service or composition of web services face a lot of problems regarding its discovery. Discovery of suitable web services has become a challenging issue due to the increasing number of selection parameters and constraints. This paper deals with a hybrid multi-agent based web service discovery mechanism which involves an artificial intelligence approach to efficiently interpret the user requirements based on both functional and non-functional demands and fuzzy constraints. The proposed model utilizes the services of intelligent software agents. Some of the agents like the reputation agent analyses the popularity of web services and assign ranks to the web services based on user feedback and statistical information. The behavior of individual user is being tracked, from which the intelligent agent interprets the fuzzy requirements of users through set of logical and analytical calculations whereas the composition agent provides flexibility to the user for custom composition of web service packages using different individual web services. In brief, we present an automated customer-centric web service discovery and composition approach which aims in an efficient web service discovery and composition followed by customer satisfaction using multi agents.