Understanding People’s Personalized and Contextual Choices of Differential Privacy: A Proof-of-Concept Survey

Abstract

Differential privacy (DP) is a state-of-the-art privacy-preserving mechanism. In this paper, we argue that interpreting users’ personalized and contextual privacy choices is key to applying DP in actual use cases. Through the lens of contextual integrity (CI), we conducted a proof-ofconcept survey (N=23) to examine how app types and information receivers affected people’s perceived appropriateness of information disclosures and their choices of the privacy-data utility tradeoff in DP. Through the exploratory analysis, we revealed people’s diverse privacy choices, which were affected by contexts. Further, people’s perceived appropriateness of information flows and desired tradeoff between privacy and data utility in DP were consistent. Based on the findings, we point out the technical questions and uncertainty about DP and stress the importance of understanding users’ personalized and contextual privacy choices to avoid misalignment between app and users and, therefore, enhance the usability of DP. This research sheds light on making DP more socially aware and adaptive to user needs via integration with the CI framework.

Keywords: Contextual Integrity, Differential Privacy