Set-Expression Based Method for Effective Privacy Preservation

Anonymization is proposed to alleviate the problem of privacy disclosure in recent years. Previous approaches typically generalize the specific values in the original data table to achieve anonymization. However, these solutions suffer from information loss in different degrees. In this paper, we propose the concept of set-expression, which causes less data distortion and equipments the anonymized table with better aggregate query answerability. Furthermore we propose a greedy algorithm and conduct a set of extensive experiments to show the advantages of our approach. We also introduce a novel Information Loss Metric to measure the quality of our proposed method.