Differentially Private Releasing via Deep Generative Model (Technical Report)

TL;DR

This paper introduces dp-GAN, a differentially private deep generative model that enables privacy-preserving release of complex data like images and text, maintaining utility and scalability.

Contribution

The paper presents a novel framework, dp-GAN, that trains generative models with differential privacy, allowing unlimited synthetic data generation for diverse analysis tasks.

Findings

Provides theoretical privacy guarantees.

Maintains high utility of generated data.

Demonstrates scalability and stability in training.

Abstract

Privacy-preserving releasing of complex data (e.g., image, text, audio) represents a long-standing challenge for the data mining research community. Due to rich semantics of the data and lack of a priori knowledge about the analysis task, excessive sanitization is often necessary to ensure privacy, leading to significant loss of the data utility. In this paper, we present dp-GAN, a general private releasing framework for semantic-rich data. Instead of sanitizing and then releasing the data, the data curator publishes a deep generative model which is trained using the original data in a differentially private manner; with the generative model, the analyst is able to produce an unlimited amount of synthetic data for arbitrary analysis tasks. In contrast of alternative solutions, dp-GAN highlights a set of key features: (i) it provides theoretical privacy guarantee via enforcing the differential privacy principle; (ii) it retains desirable utility in the released model, enabling a variety of otherwise impossible analyses; and (iii) most importantly, it achieves practical training scalability and stability by employing multi-fold optimization strategies. Through extensive empirical evaluation on benchmark datasets and analyses, we validate the efficacy of dp-GAN.