K-anonymised reducts

Abstract

Privacy-preserving data mining aims to prevent the exposure of sensitive information as a result of mining algorithms. This is commonly achieved by data anonymisation. One way to anonymise data is by adherence to the k-anonymity concept which requires that the probability to identify an individual by linking databases does not exceed 1/k. In this paper, we propose an algorithm which utilises rough set theory to achieve k-anonymity. The basic idea is to partition the original dataset into several disjoint reducts such that each one of them adheres to k-anonymity. We show that it is easier to make each reduct comply with k-anonymity if it does not contain all quasi-identifier attributes. Moreover, our procedure ensures that even if the attacker attempts to rejoin the reducts, the k-anonymity is still preserved. Unlike other algorithms that achieve k-anonymity, the proposed method requires no prior knowledge of the domain hierarchy taxonomy.