摘要
网络社交中数据传输便捷性、共享性,增加了隐私信息保护的困难性,差分隐私技术因能够低于攻击者全部背景知识,而备受关注。为此,文章针对非交互式的场景下,基于隐私保护及数据可用性的双重需求,针对数据的属性分类,分别引入Laplace机制及指数机制,通过噪声添加实现差分隐私保护;并基于数据可用性,利用改进的LWSPA查询结构序列的分割,降低向量温度,控制发布数据集的误差,以优化数据可用性,以此建构的差分隐私保护机制,能够实现隐私保护与数据可用的均衡发展。
Data transmission convenience and sharing increase the difficulty of privacy information protection.The differential privacy technology attracts much attention because it is lower than the full background knowledge of the attacker.Therefore,based on the dual demand of privacy protection and data availability,we introduce Laplace mechanism and index mechanism,realize differential privacy protection through noise addition,and use the segmentation of improved LWSPA,query structure sequence,reduce the vector temperature,control the error of the release data set,and optimize the data availability to realize the balanced development of privacy protection and data availability.
作者
杨文娟
Yang Wenjuan(Shanghai Zhongqiao Vocational And Technical University,Shanghai 201514,China)
出处
《无线互联科技》
2021年第22期28-30,共3页
Wireless Internet Technology
