Symmetric-Key Based Privacy-Preserving Scheme For Mining Support Counts

摘要

In this paper we study the problem of mining support counts using symmetric-key crypto which is more efficient than previous work. Consider a scenario that each user has an option (like or unlike) of the specified product, and a third party wants to obtain the popularity of this product. We design a much more efficient privacy-preserving scheme for users to prevent the loss of the personal interests. Unlike most previous works, we do not use any exponential or modular algorithms, but we provide a symmetric-key based method which can also protect the information. Specifically, our protocol uses a third party that generates a number of matrixes as each user’s key. Then user uses these key to encrypt their data which is more efficient to obtain the support counts of a given pattern.