Bayesian Adversarial Privacy

TL;DR

This paper introduces a new Bayesian-based privacy measure that is more contextually meaningful than differential privacy, emphasizing decision-making from the prior viewpoint and providing detailed examples and computational methods.

Contribution

It proposes a novel, context-aware privacy definition grounded in Bayesian decision theory, offering a more explicit alternative to existing frameworks.

Findings

The new privacy measure is more meaningful than differential privacy.

The approach emphasizes prior-based decision making for disclosures.

Computational methods and toy examples illustrate the framework.

Abstract

Theoretical and applied research into privacy encompasses an incredibly broad swathe of differing approaches, emphasis and aims. This work introduces a new quantitative notion of privacy that is both contextual and specific. We argue that it provides a more meaningful notion of privacy than the widely utilised framework of differential privacy and a more explicit and rigorous formulation than what is commonly used in statistical disclosure theory. Our definition relies on concepts inherent to standard Bayesian decision theory, while departing from it in several important respects. In particular, the party controlling the release of sensitive information should make disclosure decisions from the prior viewpoint, rather than conditional on the data, even when the data is itself observed. Illuminating toy examples and computational methods are discussed in high detail in order to highlight the specificities of the method.