Recommender algorithms are widely used, ranging from traditional Video on Demand to a wide variety of Web 2.0 services. Unfortunately, the related privacy concerns have not received much attention. In this paper, we study the privacy concerns associated with recommender algorithms and present a cryptographic security model to formulate the privacy properties. We propose two privacy-preserving content-based recommender algorithms and prove their properties. Moreover, we show the potential weakness in some existing collaborative filtering algorithms which claim to provide privacy protection.
|Publisher||IEEE Computer Society|
|Conference||International Conference on Collaboration Technologies and Systems, CTS 2012|
|Period||21/05/12 → 25/05/12|
|Other||21-25 May 2012|
- Recommender algorithms