D2P: Distance-Based Differential Privacy in Recommenders

The upsurge in the number of web users over the last two decades has resulted in a significant growth of online information. This information growth calls for recommenders that personalize the in- formation proposed to each individual user. Nevertheless, person- alization also opens major privacy concerns. This paper presents D2P , a novel protocol that ensures a strong form of differential privacy, which we call distance-based differen- tial privacy, and which is particularly well suited to recommenders. D2P avoids revealing exact user profiles by creating altered pro- files where each item is replaced with another one at some distance . We evaluate D 2 P analytically and experimentally on MovieLens and Jester datasets and compare it with other private and non-private recommenders.


Published in:
VLDB Endowment, 8, 862-873
Year:
2015
Publisher:
New York, Assoc Computing Machinery
Keywords:
Laboratories:


Note: The status of this file is: EPFL only


 Record created 2015-05-28, last modified 2018-12-03

Preprint:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)