基于简单随机抽样的大数据可信性验证方法

摘要

In this paper,we propose an approach to verify the trustworthiness of big data including the trustworthiness of data source and data content.The trustworthiness of data source is achieved by verifying the identity of data owner based on public key certificate.We design an interactive challenge-response protocol to verify the trustworthiness of data content on the basis of simple random sampling test and aggregated signature-based broadcast encrypting scheme.The protocol can make the users who know the public parameters of the data owner validate the verification result with high probability by sampling only a small amount of data.Formal security analysis and experimental evaluations are also conducted,showing that the proposed scheme is practical to achieve the design goal for big data.%针对大数据的可信性验证问题,本文提出了一种大数据可信性验证方法以验证数据来源和数据内容的可信性.本文首先通过验证数据属主的身份证书来实现数据来源可信性的验证,再在简单随机抽样和可聚合的广播签名方案的基础上,设计了一个交互式质询-应答协议,使得用户只需抽样少量数据就能以高置信率验证数据内容的可信性.理论分析和实验结果表明,本文方法是安全的,且性能开销在合理范围内,能够实现大数据的可信性验证.