基于熵分类的个性化隐私匿名方法

Entropy-based classification approach for personalized privacy anonymity

针对传统(α,k)-匿名模型不能满足敏感属性值之间不同隐私保护程度个性化需求的问题,引入敏感属性值个性隐私敏感因子和个性隐私保护需求度的概念,进而形式化地定义了个性化(α,k)-匿名模型;同时,还提出了一种基于熵分类的个性化隐私匿名方法来实现个性化(α,k)-匿名模型。实验表明:该方法不仅能获得与现有(α,k)-匿名算法近似的信息损失度和时间代价,同时也满足了个性化服务的需求,获得更合理的隐私保护。

To solve the defects of the traditional （α,κ）-anonymity model has the defect that it does not satisfy the personalized requirement of different privacy preserving degree for different sensitive attributed values. To overcome this defect, the concepts of personalized privacy sensitive factor and personalized privacy preserving requirement degrees for each sensitive attribute value are introduced. Then the personalized （α,κ）-anonymity model is defined formally. Meanwhile, an entropy-based classification approach for personalized privacy anonymity is presented to solve this personalized （α,κ）-anonymity model. Experiment results show that the proposed method not only produces similar information loss and time cost to the existing （α,κ）-anonymity algorithm, but also meets the requirements of personalized service and achieves more reasonable privacy preservation.