An Adaptive Evaluation Model of Web Service Based on Artificial Immune Network

There will be many Web services in the future, which have the similar or same functions, the optimization selection for these Web services is difficult for user. To solve the problem, this paper presents an adaptive evaluation model of Web service based on artificial immune network. Evaluation tree is established that is adaptive for different services. According leaves of evaluation tree, vectors of Qos attributes are acquired form performance monitoring module. AiNet immune algorithm with new mechanism of antibody promotion and suppression is used by which vectors of Qos attributes are clustered. Using a cluster with maximum sample number, average value of Qos is gotten. According to average value Web service level is defined. When values of Qos outweigh average value, these Web service will be considered to have high quality, and they can satisfy the needs of users. It is proved by experiment results that the model is effective and able to overcome the localization of the existing methods which are only based on function-optimized selection for Web services.