Abstract: With the rapid development of service computing, how to quickly and accurately find the best optimal Web service composition is the most important one of many challenges. This paper presents a second-order Hidden Markov Model（HMM）-based Web service selection method. This method uses Quality of Service（QoS） parameters to distinguish the Web services with the same functionality, and selects a group of the optimal Web services to perform user requests. By considering the two QoS parameters - throughput and response time, the method can solve the problem of measuring the quality of Web services based on thresholds set. A set of optimal services are selected to meet the needs of users by this model and algorithm. The simulation experiments validates this method is effective.

Key words: second-order hidden Markov model, Quality of Service （QoS）, Web service selection

摘要： 随着服务计算的快速发展，如何快速而准确地找到最优的Web服务组合是众多挑战中最重要的一项。提出了一种基于二阶隐马尔可夫模型（HMM）的服务选择方法。该方法使用服务质量（QoS）参数去区分具有相同功能的Web服务，并且选择一组最优的Web服务来执行用户请求。通过考虑两个QoS参数-吞吐量和响应时间，该方法能够解决根据设定的阈值来衡量Web服务质量的问题。通过构建的模型和算法，方法能够选择出最优的Web服务以满足用户的需求。仿真实验验证了所提出的方法是有效的。

关键词: 二阶隐马尔可夫模型, 服务质量（QoS）, Web服务选择