Document Type

Conference Paper

Department/Unit

Department of Computer Science

Title

Thwarting passive privacy attacks in collaborative filtering

Language

English

Abstract

While recommender systems based on collaborative filtering have become an essential tool to help users access items of interest, it has been indicated that collaborative filtering enables an adversary to perform passive privacy attacks, a type of the most damaging and easy-to-perform privacy attacks. In a passive privacy attack, the dynamic nature of a recommender system allows an adversary with a moderate amount of background knowledge to infer a user's transaction through temporal changes in the public related-item lists (RILs). Unlike the traditional solutions that manipulate the underlying user-item rating matrix, in this paper, we respond to passive privacy attacks by directly anonymizing the RILs, which are the real outputs rendered to an adversary. This fundamental switch allows us to provide a novel rigorous inference-proof privacy guarantee, known as δ-bound, with desirable data utility and scalability. We propose anonymization algorithms based on suppression and a novel mechanism, permutation, tailored to our problem. Experiments on real-life data demonstrate that our solutions are both effective and efficient. © 2014 Springer International Publishing Switzerland.

Publication Date

2014

Source Publication Title

Database Systems for Advanced Applications: 19th International Conference, DASFAA 2014, Bali, Indonesia, April 21-24, 2014. Proceedings, Part II

Start Page

218

End Page

233

Conference Location

Bali, Indonesia

Publisher

Springer International Publishing

DOI

10.1007/978-3-319-05813-9_15

ISBN (print)

9783319058122

ISBN (electronic)

9783319058139

This document is currently not available here.

Share

COinS