大数据环境下,数据具有种类多、数量大、增长速度快及价值密度低等特点,若对所有的隐私数据都提供相同程度的保护必然会造成计算资源的浪费,因此必须对隐私数据施行分级保护.差分隐私是具有严格数学定义的隐私保护模型,其以概率为基础量化了隐私保护程度,可以利用隐私预算ε对隐私保护程度划分等级.假设存在隐私保护等级的前提下,提出了分级隐私保护服务模型,基于VCG机制与最优匹配相结合的方法,为各级隐私保护服务制定合理的价格以引导用户理性地选择隐私保护服务等级.运用该机制为6个等级的隐私保护服务制定了相应的价格.分析表明,该服务模型中的定价机制可以合理地制定每个等级之间的价格,实现了隐私数据分级保护,优化了社会资源的配置.%Data has many characteristics like various kinds,large quantity,high increment speed and low value density under the environment of big data.Meanwhile,computing resources can be wasted if all the privacy data be protected by the same kind of protection level.Therefore,privacy data must be protected in a hierarchical way.Since differential privacy is a strictly mathematical defined privacy preserve model which quantifies the level of privacy preserve based on probability theory,its parameter ε can be used to rank the levels of privacy preserve.A hierarchical privacy preserve service model has been proposed under the hypothesis that there already exists certain rank of the privacy preserve levels.The hierarchical service pricing mechanism in the model is composed of VCG mechanism and optimal matching theory,which make reasonable prices of every rank of services for users to choose reasonably.The prices for six levels of privacy preserve service have been established with this proposed mechanism.The results of analysis show that the pricing mechanism in the model proposed has widen the gap among all the six levels to fulfill the hierarchical privacy preserve service and to balance the allocation of social resources.
展开▼
