基于滑动窗口的敏感关联规则隐藏

Sensitive association rule hiding based on sliding window

提出了一种新的基于数据流的关联规则隐藏算法HSRDS,采用一种变化的频繁模式树(TFP-Tree)作为原始数据与清洗后数据的过渡结构,并结合滑动窗口技术,快速实现最新数据的敏感规则隐藏。同时在进行敏感关联规则隐藏的过程中,提出了两个阈值σ和δ,使得保证敏感规则成功隐藏的同时,也能最小化隐藏规则所产生的负面效应。

We propose an algorithm, HSRDS, which achieves the sensitive association rule hiding based on data stream. In this algorithm, we adopt Transformative FP-Tree （TFP-Tree） as transitional structure between original data set and sanitized data set, and complete the sensitive association rule hiding of the newest data combined with sliding window. Furthermore, in the progress of hiding sensitive association rule, we inject two thresholds that are σ and δ, to guarantee that the sensitive association rule can be hidden successfully, at the same time the side effect, resulting from the hiding sensitive rule, can be minimized.