DP2-Pub: Differentially Private High-Dimensional Data Publication with Invariant Post Randomization

TL;DR

This paper introduces DP2-Pub, a two-phase differentially private mechanism for publishing high-dimensional data that maintains data utility through attribute clustering and invariant post randomization, suitable for real-world applications.

Contribution

The paper proposes a novel two-phase mechanism combining attribute clustering and invariant PRAM to improve utility in differentially private high-dimensional data publication.

Findings

Significantly improves data utility while satisfying differential privacy.

Effective in high-dimensional, heterogeneous data scenarios.

Extends to semi-honest server settings with local differential privacy.

Abstract

A large amount of high-dimensional and heterogeneous data appear in practical applications, which are often published to third parties for data analysis, recommendations, targeted advertising, and reliable predictions. However, publishing these data may disclose personal sensitive information, resulting in an increasing concern on privacy violations. Privacy-preserving data publishing has received considerable attention in recent years. Unfortunately, the differentially private publication of high dimensional data remains a challenging problem. In this paper, we propose a differentially private high-dimensional data publication mechanism (DP2-Pub) that runs in two phases: a Markov-blanket-based attribute clustering phase and an invariant post randomization (PRAM) phase. Specifically, splitting attributes into several low-dimensional clusters with high intra-cluster cohesion and low inter-cluster coupling helps obtain a reasonable allocation of privacy budget, while a double-perturbation mechanism satisfying local differential privacy facilitates an invariant PRAM to ensure no loss of statistical information and thus significantly preserves data utility. We also extend our DP2-Pub mechanism to the scenario with a semi-honest server which satisfies local differential privacy. We conduct extensive experiments on four real-world datasets and the experimental results demonstrate that our mechanism can significantly improve the data utility of the published data while satisfying differential privacy.