A Privacy-Preserving Architecture for the Semantic Web Based on Tag Suppression

Abstract

We propose an architecture that preserves user privacy in the semantic Web via tag suppression. In tag suppression, users may wish to tag some resources and refrain from tagging some others in order to hinder privacy attackers in their efforts to profile users’ interests. Following this strategy, our architecture helps users decide which tags should be suppressed. We describe the implementation details of the proposed architecture and provide further insight into the modeling of profiles. In addition, we present a mathematical formulation of the optimal tradeoff between privacy and tag suppression rate.