An enhanced l-diversity privacy preservation

Abstract

As a serious concern in data publishing and analysis, privacy preservation of individuals has received much attentions. Anonymity models via generalization can protect individual privacy, but often lead to superabundance information loss. Therefore, privacy preserving data publishing needs a careful balance between privacy protection and data utility. The challenge is how to lessen the information loss during anonymity. This paper presents a (k, l, θ)-diversity model base on clustering to minimize the information loss as well as assure data quality. We take into accounts the cluster size, the distinct sensitive attribute values and the privacy preserving degree for this model. Besides, we account for the complexity of our algorithm. Extensive experimental evaluation shows that our techniques clearly outperform the existing approaches in terms of execution time and data utility.