基于属性熵的隐私匿名信息保护研究与应用

Privacy Preserving Anonymous Information Protection Based on Attribute Entropy

针对当前隐私匿名保护中存在的只对准标识符属性与敏感属性进行保护,而未考虑两者之间的平衡性问题,结合当前的智能算法,提出了一种基于权重属性熵的匿名分类方法。首先提出了方法的思路,然后引入权重属性熵的方式,对元素损失度量模型和分类匿名数据保护模型各自所占的权重进行计算,进行找到损失度量和匿名保护之间的平衡点。最后通过试验验证的方式,验证了本文提出的算法虽然执行时间多,但是可实现数据匿名保护和精度之间的平衡,具有一定的可行性。

In order to balance the two attributes, an anonymous classification method is proposed based on weighted attribute entropy. The method tries to solve the problem existed in the current methods that aim at protecting only the attributes of identifiers or sensitive attributes. Firstly, the idea of the method is put forward. Then, the weights of element loss measurement model and classified anonymous data protection model are calculated by introducing weight attribute entropy, and the balance point between loss measurement and anonymous protection is found. Finally, the experimental verification shows that the proposed algorithm can achieve a balance between data anonymity protection and accuracy, although it takes a lot of time to execute, it has a certain feasibility.