【作者】 石磊；

【导师】 刘振鹏；

【作者基本信息】 河北大学 ， 计算机应用技术， 2006， 硕士

【摘要】 P2P环境因其分布式、易拓展等优势，正在赢得广泛的关注，越来越多的Web服务被部署到上面。Web服务中的关键性的问题是服务选择，最终的用户必须选择一个好的服务提供者。由于P2P环境的特性，往往每次查询会返回大量的结果，如何在结果中进行选择，成为了关键性的问题。 本文基于一种广为应用的高效的P2P路由算法．分布式哈希表算法，构建了P2P环境中的服务推举网络，并给出了服务选择的算法和框架。其流程是：发出查询，收集答案(包括推举)，进行本地比较，选择最优服务，并根据选择的服务，对其进行评估，然后调整相应的模型，这些模型包括服务质量向量和服务推举可信度。最后通过实验对系统的精确性以及动态可调整性进行了验证。更多还原

【Abstract】 P2P is increasingly receiving attention in research because of its distributed and scalability, and more and more Web services are deployed on it. However, the key challenge is Web service selection: ultimately, the service user must select a good Web service provider. The query for Web service in P2P often lead to more answers, so it become the key challenge about how to select the good answer.In this paper, we construct a scalable distributed referral network in P2P and propose a heuristic algorithm for Web services selection through distributed hash table method, which is the most popular and efficient P2P routing algorithm. The work flow include: send out query, collect answers (include referrals), compute locally, select the best answer according to the computing result, rate the provider and modify the models saved in the requester, the model include the quality vector and the referral credibleness. Finally, the experiment results show the accuracy and efficiency of our framework.更多还原