电子交易中保护消费者隐私相关性分析

Correlation Analysis on Consumer Privacy-protecting Big-data in Electronic Transaction Systems

对电子交易中的消费数据进行统计分析能方便向消费者提供更好的服务.针对消费数据容易暴露消费者隐私等安全问题,提出了一种有效实用的解决方案,首先将消费者数据分块存储在多服务器中,同时服务器采用Paillier同态加密算法对各自所存储的数据加密,并对分析者进行访问控制,确保仅授权分析者可获得结果.最后在密文域上完成安全计算与统计分析,有效防止半诚实服务器的内部攻击.分析表明,在不泄露消费者隐私的前提下,可有效完成相关统计分析,具有较高实用价值.

In order to protect consumer data from attacks in electronic transactions,a practical privacy-preserving correlation analysis scheme was proposed.To prevent internal attacks,consumer data were separated into several parts and stored in different servers.The access was controlled to ensure that the unauthorized analyzer could not obtain any results.The homomorphic Paillier cryptosystem was used for the security calculation.To ensuring consumer privacy,the statistical correlation analysis of the big data were implemented.The correctness of the scheme was analyzed,and the computational and communication complexity were also provided,which indicated the scheme could achieve the efficient analysis without revealing consumers′privacy.