Adaptive Web Services Composition for Multi-user

In a composite service for multiple users, network position of multiple users are different, leading to users perceptions to one service are different. In addition, network parameter values, which vary due to many reasons (such as network congestions), influence QoS of Web services. We propose a reliable service composition method which uses multiple candidate services to process service requests for a variety of users. Initially, multiple service execution paths for each kind of user requests are calculated. In runtime, the optimal services are chosen according to the current network state. We present an application scenarios for the method.