A Second-Order Hidden Markov Model Based Web Services Selection
Yuan Lu, Zhichun Jia, Xiang Li, Xing Xing · 2018 IEEE 7th Data Driven Control and Learning Systems Conference (DDCLS) · 2018
Over the last few decades, the cloud computing is rapidly developing. How to find the suitable web services for users quickly and accurately is facing more and more challenges. The quality of service becomes an essential parameter to discriminate web services with the same function. In this paper, we propose an effective services selection method based on Quality of Service (QoS) parameters. Our method uses the second-order Hidden Markov Model (HMM) to model the business process of web services and selects the optimal web services for the execution of user requests. The technique we present can solve the measurement problem of the web service behaviors according to the given threshold values of the throughput and response time. By ranking the candidate services with the similar functionality, the top service is selected to run in the business process for meeting the user needs. Finally, we conduct the simulation experiments to demonstrate our method using QWS database. The result shows that our method is effective.