Privacy risks in recommender systems

Naren Ramakrishnan, Benjamin J. Keller, Batul J Mirza, Ananth Y Grama, George Karypis

Abstract

The authors explore the conflict between personalization and privacy that
arises from the existence of straddlers-users with eclectic tastes who rates products across several different types or domains--in recommender systems. While straddlers enable serendipitous recommendations, information about their existence could be used in conjunction with other data sources to uncover identities and reveal personal details. This article discusses a graph? theoretic model for studying the benefit for and risk to straddlers ...