On differentially private online collaborative recommendation systems

Abstract

In collaborative recommendation systems, privacy may be compromised, as users’ opinions are used to generate recommendations for others. In this paper, we consider an online collaborative recommendation system, and we measure users’ privacy in terms of the standard notion of differential privacy. We give the first quantitative analysis of the trade-offs between recommendation quality and users’ privacy in such a system by showing a lower bound on the best achievable privacy for any algorithm with non-trivial recommendation quality, and proposing a near-optimal algorithm. From our results, we find that there is actually little trade-off between recommendation quality and privacy, as long as non-trivial recommendation quality is to be guaranteed. Our results also identify the key parameters that determine the best achievable privacy.