Randomized Response and Differential Privacy
Andreas Ioannidis, Αντώνιος Λίτκε, Nikolaos K. Papadakis · 2024
In this paper we reconsider the classical concept of Randomized Response and investigate how it relates with data privacy and especially Differential Privacy. Our purpose is to formalize Randomized Response in a natural probability theoretic setting and investigate its statistical features from a mathematical point of view. More precisely, we propose a solid probabilistic framework which can be easily generalized to more complex cases. The analysis of the proposed framework shows its flexibility to be applied as a suitable differential privacy solution of sharing medical data over distributed cloud infrastructures.