Differential Privacy Data Release Scheme Using Microaggregation With Conditional Feature Selection

Differential privacy (DP) has achieved great progress in addressing the user privacy preservation issues related to data analysis in the Internet of Things (IoT) services and applications. However, the existing DP models tend to overlook the effect of the correlation of data features on the utility of the smart IoT data. To mitigate this gap, we propose a Microaggregation-based DP method using Conditional Mutual Information (M-DPCMI) for prerelease data processing and feature selection. With the new method, we leverage the anonymous microaggregation approach to improve data utility while preventing potential sensitive IoT user information leakage. In addition, M-DPCMI is theoretically proved to satisfy the definition of DP and experimentally validated over real data sets. It is shown that the new model achieves better data utility than the state-of-the-art DP methods.