Anonymous statistical survey of attributes

Abstract

A distributor of digital contents desires to collect users’ attributes. This is because the distributor can grasp the image of users, and work out the marketing strategy. On the other hand, the users do not desire to offer the attributes owing to the privacy protection. For anonymous surveys, a protocol to generate statistical results of the attributes is previously proposed, where the extra information is not released beyond the statistical results. However, in the simple application of this protocol to the surveys, the correctness of the statistical results is not assured, since the users do not necessarily offer the correct attributes. In this paper, under the assumption that some trusted third parties exist, an anonymous statistical survey system of attributes witht he correctness is proposed.