个性化K-匿名模型

K-匿名化是数据发布环境下保护数据隐私的一种方法.目前的K-匿名化方法主要是针对一些预定义的隐私泄露参数来进行隐私控制的.隐私保护的重要原则之一就是隐私信息的拥有者有隐私自治的权利[1].这就要求在实现匿名化过程当中考虑到个人不同的隐私需求,制定个性化的隐私约束.根据个人隐私自治的原则结合K-匿名模型的最新发展,提出了一种个性化K-匿名模型,并给出了基于局部编码和敏感属性泛化的个性化K-匿名算法.实验结果表明,该方法可以在满足个性化隐私需求的情况下,完成匿名化过程,并且采用该方法进行匿名所造成的信息损失较小.

Personalized K-anonymity

K-anonymity is a popular model used in microdata publishing.Currently,K-anonymity researches focus on protecting privacy using pre-defined system parameters.One important principle in privacy preserving is that an individual has the right to decide his own privacy requirements.Thus,personalized privacy requirements should be taken into account when designing privacy protecting models.A Personalized K-anonymity model and corresponding anonymity method using local recoding and sensitive attribute generalization are introduced,and it's shown that the model can meet personal privacy requirements and the information loss during anonymizing process is low.