Efficiently Computing Private Recommendations
conference paper
Online recommender systems enable personalized service to users. The underlying collaborative filtering techniques operate on privacy sensitive user data, which could be misused if it is leaked or by the service provider him self. To protect user’s privacy, we propose to encrypt the data and generate recommendations by processing them under encryption. Thus, the service provider observes neither user preferences nor recommendations. The proposed method uses homomorphic encryption and secure multiparty computation (MPC) techniques, which introduce a significant overhead in computational complexity. The second contribution of this paper lies in minimizing this overhead by packing data. The improvements are illustrated by a complexity analysis.
TNO Identifier
463803
Source title
Advanced School of Computing and Imaging Conference - ASCI 2010, 1-3 November 2010, Veldhoven, The Netherlands