基于Beta信誉模型的虚假Web服务QoS反馈滤除算法

发布时间：2018-04-03 19:48

  本文选题：Web服务　切入点：服务选择　出处：《海军工程大学学报》2017年02期

【摘要】：针对由于Web服务推荐用户提交虚假QoS反馈而导致的QoS预测准确度下降问题,提出了一种基于Beta信誉模型的虚假QoS反馈滤除算法。该算法首先将若干个经过认证的中心用户作为初始可信用户集;然后,评估用户反馈数据与可信用户集反馈数据的偏离程度;最后,根据用户提交的偏离反馈次数对其信誉度进行评估,并将信誉度高于可信阈值的用户判定为可信用户,使可信用户集得到更新。通过循环执行以上过程,大部分虚假用户将被分离出系统。实验结果表明:该算法具有较强的虚假用户筛查能力,可有效提升QoS预测算法的抗攻击能力。
[Abstract]:Because the QoS prediction of Web service recommendation user submitted false feedback QoS caused a decline in the problem accurately, a false QoS Beta reputation model based on the feedback filtering algorithm. Firstly, a plurality of user authentication through the centre as the initial set of trusted users; then, user feedback and evaluation data set of trusted user feedback the degree of deviation from the data; finally, according to the user submitted from feedback times on the credibility evaluation, and the credibility is higher than the threshold for the user to determine the trusted trusted users to trusted users set updated. Through the cycle of the implementation of the above process, most of the false users will be isolated from the system. The experimental results show that the algorithm has false screening ability of the user, can effectively improve the QoS prediction algorithm anti attack capability.

【作者单位】： 海军工程大学装备经济管理系;
【基金】：国家部委基金资助项目(9140A04020215JB11050)
【分类号】：TP393.09
，

本文编号：1706653