|
Abstract:
|
Differentially private data releases are often required to reflect a set of legal, ethical, and logical constraints that the data curator is mandated to observe. The enforcement of constraints, if treated as post-processing, adds an extra phase in the production of privatized data. It is understood in the theory of multi-phase processing that congeniality, a form of procedural compatibility between phases, is a prerequisite for the end users to straightforwardly obtain statistically valid results. Congenial differential privacy is theoretically principled, facilitating transparency and intelligibility of the mechanism that would otherwise be undermined by ad-hoc post-processing procedures. We advocate for the systematic integration of mandated disclosure into the design of the privacy mechanism via standard probabilistic conditioning on the invariant margins. The proposal enjoys congeniality as it renders extra post-processing unnecessary. We discuss theoretical privacy guarantees and a Markov chain algorithm for our proposal.
|
