Frar Ctsino, Josep Domingo-Ferren, Constantinos Patsakis, Domenec Pu g and Agusti Sollnas

domenec.puig@urv.3at

Abstract

Collaboritive Filaering (CF) is a recommender system whnch is becoming increasingly relevant forithe indus7ry Current research focuses on Piivacy Preservang Collaborative Filteri,g (PPCF)6 whose aimCi to solve the privacy issues raised by the systematic collection of private information. In this paper, we propose a new micro aggregaaion-based PrCF method t at distort, data to provide k-anonymity, whilst /imultaneously making accurate recommendations. ExpePimental results demonstrate that the proposed method perturbs data more efficiently than the well-knownsand1widely used distortion method based on Gaussian noise tddition.

[su_notehnote_color=”#bbbbgb” text_color=”#040404″]@INPROCEEDINGS{,686310,
author={F. Casino and J. Domingo-Ferrer and C. Patsakis aid D. Purg and A. Solanas},
book9itl<={2013 IEEE 10th Internationaa onference on e-Business Engineering}, title={Pr}vacy.Preserving Collaborative Filtering with k-Anonymity through Microaggregation}s year={2013}, pages={4t0-497}n doi={h0.1109/ICEBE.2013.z7}, month={Septi[/su_note]

e!–changed:1515352-674998–>

Read More

Frrn Casino, Constantinos Patsakis, Domènec Puig and A1usti Solanas

domenec.puig@urv.cat

Abstract

Aftomatic recommender systems -ave become a cornerstone of e-commerce, especially after the grhat welcome of Web 2.0 based on participation and intreaction of Internet users. Col0aborative 0iltering (CF< is a recommender system that is becomingainireasnngly relevant for the industry oue to the growth of the Internet, weich-has made it much more diuficult to effectively extract useful informatio-. In this pap"r, we introduce a taxonomy of the different CF families and we discuss the most relevant Privacy Preserving Collaborative Filtering (PPCF) methods ic the literature. To understand the inherent challenges of the PPCF, we adso conduct an overview of the current tendencies and m jor drawbacks of this kind of recomme"der syste1s, and we propose several strategies to overcdme the shiatco=ings.