Multi-user Pufferfish Privacy
For privacy researchers, this provides a theoretical framework for individual indistinguishability in dynamic multi-user settings, but the contribution is incremental as it applies existing methods (Kantorovich) to a new problem.
This paper extends pufferfish privacy to multi-user systems, deriving sufficient conditions for Laplace noise calibration to achieve indistinguishability under user data changes, replacements, and additions/removals, with relaxed conditions for binary variables that reduce noise and improve utility.
This paper studies how to achieve individual indistinguishability by pufferfish privacy in aggregated query to a multi-user system. It is assumed that each user reports realization of a random variable. We study how to calibrate Laplace noise, added to the query answer, to attain pufferfish privacy when user changes his/her reported data value, leaves the system and is replaced by another use with different randomness. Sufficient conditions are derived for all scenarios for attaining statistical indistinguishability on four sets of secret pairs. They are derived using the existing Kantorovich method (Wasserstain metric of order $1$). These results can be applied to attain indistinguishability when a certain class of users is added or removed from a tabular data. It is revealed that attaining indifference in individual's data is conditioned on the statistics of this user only. For binary (Bernoulli distributed) random variables, the derived sufficient conditions can be further relaxed to reduce the noise and improve data utility.