Context-Aware Generative Adversarial Privacy

TL;DR

This paper introduces generative adversarial privacy (GAP), a novel framework using GANs to learn privacy-preserving data sanitization schemes directly from datasets, balancing privacy and utility without prior statistical knowledge.

Contribution

The paper proposes a new context-aware privacy framework that leverages GANs to learn privacy mechanisms from data, overcoming limitations of existing methods requiring dataset statistics.

Findings

GAP learns privacy mechanisms matching theoretical optima in simple models.

The framework effectively balances privacy and utility without dataset statistics.

Demonstrates practical applicability of GAN-based privacy learning.

Abstract

Preserving the utility of published datasets while simultaneously providing provable privacy guarantees is a well-known challenge. On the one hand, context-free privacy solutions, such as differential privacy, provide strong privacy guarantees, but often lead to a significant reduction in utility. On the other hand, context-aware privacy solutions, such as information theoretic privacy, achieve an improved privacy-utility tradeoff, but assume that the data holder has access to dataset statistics. We circumvent these limitations by introducing a novel context-aware privacy framework called generative adversarial privacy (GAP). GAP leverages recent advancements in generative adversarial networks (GANs) to allow the data holder to learn privatization schemes from the dataset itself. Under GAP, learning the privacy mechanism is formulated as a constrained minimax game between two players: a privatizer that sanitizes the dataset in a way that limits the risk of inference attacks on the individuals' private variables, and an adversary that tries to infer the private variables from the sanitized dataset. To evaluate GAP's performance, we investigate two simple (yet canonical) statistical dataset models: (a) the binary data model, and (b) the binary Gaussian mixture model. For both models, we derive game-theoretically optimal minimax privacy mechanisms, and show that the privacy mechanisms learned from data (in a generative adversarial fashion) match the theoretically optimal ones. This demonstrates that our framework can be easily applied in practice, even in the absence of dataset statistics.