Zero Knowledge Bi-Party Computation Using Oblivious Transfers for Recommender Systems

Abstract

Behavioural recommendations, utilizing data collected from online activity, is invaluable in helping advertisers reach the most suitable audience for their products. However, user privacy could be better pro-tected if it could be accomplished without collecting a particular individual’s data. We provide a generic construction to leverage the clas-sical Zero-Knowledge (ZK) protocol into a composable oblivious transfer (OT) protocol, preserving the ZK protocol’s round-complexity properties and security guar-antees in the resulting OT protocol. Incorporating Fuzzy Logic, this amalgamation ensures that users’ sensitive, personal and private data is always kept secure when communicating with the recommender systems while not compromising the capacity of these systems to cre-ate cohorts of similar individuals.