云环境下融合恶意用户过滤机制的信誉评估方法

发布时间：2019-07-01 14:53

【摘要】：针对云服务供应商和云用户实体双方交互过程中提供虚假或恶意信息的问题,提出了一种融合恶意用户过滤机制的信誉评估方法.首先,运用统计过程控制理论中改进的指数加权滑动平均方法对目标云服务有过反馈评级的用户进行检测,并过滤恶意用户;然后,在目标云服务的信誉计算过程中,利用多数共识理论和反馈相似度确立良性用户反馈评分的聚合权重,以提高信誉计算的准确性;最后,用户根据云服务供应商的服务和信誉情况与其进行协商,屏蔽掉信誉较低的云服务,从而确定最后的交互对象.实验结果表明,本文提出的方法不仅能够有效防止云用户的欺诈行为,而且能够杜绝云供应商的不诚信行为,使服务中心能够将信誉好的云服务推荐给用户.
[Abstract]:In order to solve the problem of providing false or malicious information between cloud service providers and cloud user entities, a reputation evaluation method based on malicious user filtering mechanism is proposed. Firstly, the improved exponential weighted moving average method in statistical process control theory is used to detect the users with feedback rating of the target cloud service and filter malicious users. Then, in the process of calculating the reputation of the target cloud service, the aggregation weight of the benign user feedback score is established by using the majority consensus theory and feedback similarity, in order to improve the accuracy of the reputation calculation. Finally, according to the service and reputation of the cloud service provider, the user negotiates with the cloud service provider to block out the cloud service with low reputation, so as to determine the final interaction object. The experimental results show that the method proposed in this paper can not only effectively prevent the fraud of cloud users, but also put an end to the dishonest behavior of cloud suppliers, so that the service center can recommend reputable cloud services to users.
【作者单位】： 燕山大学信息科学与工程学院;河北省计算机虚拟技术与系统集成重点实验室;
【基金】：国家自然科学基金项目(61379116)资助 河北省自然科学基金项目(F2015203046)资助 河北省高等学校科学技术研究重点项目(ZH2012028)资助
【分类号】：TP393.09
，

本文编号：2508575