【摘要】 针对不同敏感值的隐私保护程度需求,提出一种敏感度计算方法,将敏感值进行等级划分,再对不同等级的敏感值设定不同的敏感度;给出一种隐私保护原则(ε,k)-sensitivity来控制等价类中敏感度的分布情况,使得等价类中高敏感度的元组不会过多而造成隐私泄露;提出一种最小信息损失增量优先算法(minimum information loss increment first,MILIF)来实现隐私保护的要求。研究结果表明:所提出的方法在降低少量时间和保持数据效用的前提下,充分提高了数据表抵御敏感性攻击的能力。更多还原

【Abstract】 In order to satisfy the different privacy protection requirements for different sensitive values, a method was proposed to calculate the sensitivity of sensitive value, which was divided into several levels with different sensitivities. A(ε, k)-sensitivity principle was proposed to control the distributions of sensitivity in equivalence class and the number of the high sensitivity tuples. A minimum information loss increment first algorithm was proposed. The results show that the proposed method can improve the ability of resisting sensitivity attack, on the premise of expending a little time and maintaining a high data utility.更多还原