利用PCA增强随机化隐私数据保护方法

USING PCA TO IMPROVE RANDOMIZATION SCHEME OF PPDM

基于随机化的数据扰乱及重构技术是数据挖掘中的隐私保护(Privacy-Preserving Data Mining,PPDM)领域中最重要的方法之一。但是,随机化难以消除由于属性变量本身相关性引起的数据泄漏。介绍了一种利用主成分分析(Principal Component Anal-ysis,PCA)进行属性精简的增强随机化方法,降低了参与数据挖掘的属性数据间相关性,更好地保护了隐私数据。

Randomization, as one of the most important schemes in Privacy-Preserving Data Mining （PPDM） field, can＇t eliminate privacy breaches of datasets with high correlated attributes effectively. An improvement on randomization scheme is made through the Principal Component Analysis（PCA） to reduce the correlation between the attributes involved in data mining and preserve privacy of original data better.