A privacy-preserving framework for personalized, social recommendations

Zach Jorgensen, Ting Yu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

17 Citations (Scopus)

Abstract

We consider the problem of producing item recommendations that are personalized based on a user's social network, while simultaneously preventing the disclosure of sensitive user-item preferences (e.g., product purchases, ad clicks, web browsing history, etc.). Our main contribution is a privacypreserving framework for a class of social recommendation algorithms that provides strong, formal privacy guarantees under the model of differential privacy. Existing mechanisms for achieving differential privacy lead to an unacceptable loss of utility when applied to the social recommendation problem. To address this, the proposed framework incorporates a clustering procedure that groups users according to the natural community structure of the social network and significantly reduces the amount of noise required to satisfy differential privacy. Although this reduction in noise comes at the cost of some approximation error, we show that the benefits of the former significantly outweigh the latter. We explore the privacy-utility trade-off for several different instantiations of the proposed framework on two real-world data sets and show that useful social recommendations can be produced without sacrificing privacy. We also experimentally compare the proposed framework with several existing differential privacy mechanisms and show that the proposed framework significantly outperforms all of them in this setting.

Original languageEnglish
Title of host publicationAdvances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings
PublisherOpenProceedings.org, University of Konstanz, University Library
Pages571-582
Number of pages12
ISBN (Electronic)9783893180653
DOIs
Publication statusPublished - 2014
Event17th International Conference on Extending Database Technology, EDBT 2014 - Athens, Greece
Duration: 24 Mar 201428 Mar 2014

Other

Other17th International Conference on Extending Database Technology, EDBT 2014
CountryGreece
CityAthens
Period24/3/1428/3/14

Fingerprint

History

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems
  • Software

Cite this

Jorgensen, Z., & Yu, T. (2014). A privacy-preserving framework for personalized, social recommendations. In Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings (pp. 571-582). OpenProceedings.org, University of Konstanz, University Library. https://doi.org/10.5441/002/edbt.2014.51

A privacy-preserving framework for personalized, social recommendations. / Jorgensen, Zach; Yu, Ting.

Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings. OpenProceedings.org, University of Konstanz, University Library, 2014. p. 571-582.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Jorgensen, Z & Yu, T 2014, A privacy-preserving framework for personalized, social recommendations. in Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings. OpenProceedings.org, University of Konstanz, University Library, pp. 571-582, 17th International Conference on Extending Database Technology, EDBT 2014, Athens, Greece, 24/3/14. https://doi.org/10.5441/002/edbt.2014.51
Jorgensen Z, Yu T. A privacy-preserving framework for personalized, social recommendations. In Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings. OpenProceedings.org, University of Konstanz, University Library. 2014. p. 571-582 https://doi.org/10.5441/002/edbt.2014.51
Jorgensen, Zach ; Yu, Ting. / A privacy-preserving framework for personalized, social recommendations. Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings. OpenProceedings.org, University of Konstanz, University Library, 2014. pp. 571-582
@inproceedings{6e3d27ae3bc14db1a972af0cad585efa,
title = "A privacy-preserving framework for personalized, social recommendations",
abstract = "We consider the problem of producing item recommendations that are personalized based on a user's social network, while simultaneously preventing the disclosure of sensitive user-item preferences (e.g., product purchases, ad clicks, web browsing history, etc.). Our main contribution is a privacypreserving framework for a class of social recommendation algorithms that provides strong, formal privacy guarantees under the model of differential privacy. Existing mechanisms for achieving differential privacy lead to an unacceptable loss of utility when applied to the social recommendation problem. To address this, the proposed framework incorporates a clustering procedure that groups users according to the natural community structure of the social network and significantly reduces the amount of noise required to satisfy differential privacy. Although this reduction in noise comes at the cost of some approximation error, we show that the benefits of the former significantly outweigh the latter. We explore the privacy-utility trade-off for several different instantiations of the proposed framework on two real-world data sets and show that useful social recommendations can be produced without sacrificing privacy. We also experimentally compare the proposed framework with several existing differential privacy mechanisms and show that the proposed framework significantly outperforms all of them in this setting.",
author = "Zach Jorgensen and Ting Yu",
year = "2014",
doi = "10.5441/002/edbt.2014.51",
language = "English",
pages = "571--582",
booktitle = "Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings",
publisher = "OpenProceedings.org, University of Konstanz, University Library",

}

TY - GEN

T1 - A privacy-preserving framework for personalized, social recommendations

AU - Jorgensen, Zach

AU - Yu, Ting

PY - 2014

Y1 - 2014

N2 - We consider the problem of producing item recommendations that are personalized based on a user's social network, while simultaneously preventing the disclosure of sensitive user-item preferences (e.g., product purchases, ad clicks, web browsing history, etc.). Our main contribution is a privacypreserving framework for a class of social recommendation algorithms that provides strong, formal privacy guarantees under the model of differential privacy. Existing mechanisms for achieving differential privacy lead to an unacceptable loss of utility when applied to the social recommendation problem. To address this, the proposed framework incorporates a clustering procedure that groups users according to the natural community structure of the social network and significantly reduces the amount of noise required to satisfy differential privacy. Although this reduction in noise comes at the cost of some approximation error, we show that the benefits of the former significantly outweigh the latter. We explore the privacy-utility trade-off for several different instantiations of the proposed framework on two real-world data sets and show that useful social recommendations can be produced without sacrificing privacy. We also experimentally compare the proposed framework with several existing differential privacy mechanisms and show that the proposed framework significantly outperforms all of them in this setting.

AB - We consider the problem of producing item recommendations that are personalized based on a user's social network, while simultaneously preventing the disclosure of sensitive user-item preferences (e.g., product purchases, ad clicks, web browsing history, etc.). Our main contribution is a privacypreserving framework for a class of social recommendation algorithms that provides strong, formal privacy guarantees under the model of differential privacy. Existing mechanisms for achieving differential privacy lead to an unacceptable loss of utility when applied to the social recommendation problem. To address this, the proposed framework incorporates a clustering procedure that groups users according to the natural community structure of the social network and significantly reduces the amount of noise required to satisfy differential privacy. Although this reduction in noise comes at the cost of some approximation error, we show that the benefits of the former significantly outweigh the latter. We explore the privacy-utility trade-off for several different instantiations of the proposed framework on two real-world data sets and show that useful social recommendations can be produced without sacrificing privacy. We also experimentally compare the proposed framework with several existing differential privacy mechanisms and show that the proposed framework significantly outperforms all of them in this setting.

UR - http://www.scopus.com/inward/record.url?scp=84971306987&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84971306987&partnerID=8YFLogxK

U2 - 10.5441/002/edbt.2014.51

DO - 10.5441/002/edbt.2014.51

M3 - Conference contribution

SP - 571

EP - 582

BT - Advances in Database Technology - EDBT 2014: 17th International Conference on Extending Database Technology, Proceedings

PB - OpenProceedings.org, University of Konstanz, University Library

ER -