An Architecture for Privacy Preserving Collaborative Filtering on Web Portals

Popular E-commerce portals such as Amazon and eBay require user personal data to be stored on their servers for serving these users with personalized recommendations. These recommendations are derived by virtue of collaborative filtering. Collaborative filtering (CF) is a method to perform automated recommendations based upon the assumption that users who had similar interests in past, will have similar interests in future too. Storing user personal information at such servers has given rise to a number of privacy concerns [13] which are effecting business of these services [15]. In this paper, we present a novel architecture for privacy preserving collaborative filtering for these services. The proposed architecture attempts to restore user trust in these services by introducing the notion of ´distributed trust´. This essentially mean that instead of trusting a single server, a coalition of servers is trusted. Distributions of trust makes the proposed architecture fault resilient and robust against security attacks. Moreover, the architecture employs an efficient crossing minimization based biclustering algorithm for collaborative filtering. This algorithm is easily amenable to privacy preserving implementation. The privacy preserving implementation makes use of a threshold homomorphic cryptosystem. The proposed algorithm is fully implemented and evaluated with encouraging results.