基于DHT和VSM的Web服务推举策略

Web Services Referral Strategy Based on DHT and VSM in P2P

由于P2P环境的分布式、易拓展等优点,越来越多的Web服务被部署到上面,但是如何在大量返回的结果中进行选择,是Web服务的关键问题。基于分布式哈希表路由算法,并引入向量空间模型相关思想,构建P2P环境中的服务推举网络,提出服务选择的思想和框架。其流程是:发出查询;接收结果(包括推举);进行本地比较,选择最优服务;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。

More and more Web services are deployed on P2P because of its distributivity and scalability. However,it becomes the key challenge to select a good one from numerous answers. A scalable distributed referral network in P2P is constructed,and a heuristic algorithm for Web services selection through Distributed Hash Table method（DHT） is proposed,and Vector Space Model（VSM） is introduced. The work flow includes sending out query ,collecting answers （include referrals）, computing local services, selecting the best answer according to the computing result,rateing the provider and modifying the models saved in the requester. The model includes the quality vector and the referral reliability. The accuracy and efficiency of this framework is proved in the experiment.