Association rule mining algorithm based on privacy preserving

The goal of privacy preserving association rule is to find an approach of the original data set, making the relevant sensitive rules not to be found in a process of data mining. In order to efficiently achieve privacy preserving, this paper proposes an effective method for privacy preserving association rule. Trough the way of combining the basic strategies of PPARM algorithm and IMBA which are improved, making less-sensitive operations, a greater degree of hiding can be made while a less degree impact for the non-sensitive rules can be made too. Theoretical and experimental results show that the algorithm is highly efficient, and can ensure the privacy preserving better.

I. Introduction

In recent years, data mining based on privacy preserving has become a hot spot of database study. Association rules are researched popularly too. Association rule mining is to find the interesting relationship and interrelationship of item sets among the mass data. The privacy preserving of association rule is an operation done to hide the sensitive rules aimed to frustrate the data user's attempt to mine the data from its owner before the process of association rule mining. To implement the privacy preserving of data, we need to consider the follow two points: One is how to ensure the privacy not to be revealed during data application process. Another is how to make it better for data application. At present, the study of privacy preserving is concentrated on how to design a better algorithm while balance the two points mentioned above^[2].