A Differential Privacy Framework for Collaborative Filtering

A Differential Privacy Framework for Collaborative Filtering

Focusing on the privacy issues in recommender systems, we propose a framework containing two perturbation methods for differentially private collaborative filtering to prevent the threat of inference attacks against users. To conceal individual ratings and provide valuable predictions, we consider s...

Full description

Bibliographic Details
Main Authors: Jing Yang, Xiaoye Li, Zhenlong Sun, Jianpei Zhang
Format: Article
Language:English
Published: Hindawi Limited 2019-01-01
Series:Mathematical Problems in Engineering
Online Access:http://dx.doi.org/10.1155/2019/1460234
id doaj-d661ef31ca5c4a65b9f2ecccbcb1180b
record_format Article
spelling doaj-d661ef31ca5c4a65b9f2ecccbcb1180b2020-11-24T21:56:45ZengHindawi LimitedMathematical Problems in Engineering1024-123X1563-51472019-01-01201910.1155/2019/14602341460234A Differential Privacy Framework for Collaborative FilteringJing Yang0Xiaoye Li1Zhenlong Sun2Jianpei Zhang3College of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang 150001, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang 150001, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang 150001, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, Heilongjiang 150001, ChinaFocusing on the privacy issues in recommender systems, we propose a framework containing two perturbation methods for differentially private collaborative filtering to prevent the threat of inference attacks against users. To conceal individual ratings and provide valuable predictions, we consider some representative algorithms to calculate the predicted scores and provide specific solutions for adding Laplace noise. The DPI (Differentially Private Input) method perturbs the original ratings, which can be followed by any recommendation algorithms. By contrast, the DPM (Differentially Private Manner) method is based on the original ratings, which perturbs the measurements during implementation of the algorithms and releases the predicted scores. The experimental results showed that both methods can provide valuable prediction results while guaranteeing DP, which suggests it is a feasible solution and can be competent to make private recommendations.http://dx.doi.org/10.1155/2019/1460234
collection DOAJ
language English
format Article
sources DOAJ
author Jing Yang
Xiaoye Li
Zhenlong Sun
Jianpei Zhang
spellingShingle Jing Yang
Xiaoye Li
Zhenlong Sun
Jianpei Zhang
A Differential Privacy Framework for Collaborative Filtering
Mathematical Problems in Engineering
author_facet Jing Yang
Xiaoye Li
Zhenlong Sun
Jianpei Zhang
author_sort Jing Yang
title A Differential Privacy Framework for Collaborative Filtering
title_short A Differential Privacy Framework for Collaborative Filtering
title_full A Differential Privacy Framework for Collaborative Filtering
title_fullStr A Differential Privacy Framework for Collaborative Filtering
title_full_unstemmed A Differential Privacy Framework for Collaborative Filtering
title_sort differential privacy framework for collaborative filtering
publisher Hindawi Limited
series Mathematical Problems in Engineering
issn 1024-123X
1563-5147
publishDate 2019-01-01
description Focusing on the privacy issues in recommender systems, we propose a framework containing two perturbation methods for differentially private collaborative filtering to prevent the threat of inference attacks against users. To conceal individual ratings and provide valuable predictions, we consider some representative algorithms to calculate the predicted scores and provide specific solutions for adding Laplace noise. The DPI (Differentially Private Input) method perturbs the original ratings, which can be followed by any recommendation algorithms. By contrast, the DPM (Differentially Private Manner) method is based on the original ratings, which perturbs the measurements during implementation of the algorithms and releases the predicted scores. The experimental results showed that both methods can provide valuable prediction results while guaranteeing DP, which suggests it is a feasible solution and can be competent to make private recommendations.
url http://dx.doi.org/10.1155/2019/1460234
work_keys_str_mv AT jingyang adifferentialprivacyframeworkforcollaborativefiltering
AT xiaoyeli adifferentialprivacyframeworkforcollaborativefiltering
AT zhenlongsun adifferentialprivacyframeworkforcollaborativefiltering
AT jianpeizhang adifferentialprivacyframeworkforcollaborativefiltering
AT jingyang differentialprivacyframeworkforcollaborativefiltering
AT xiaoyeli differentialprivacyframeworkforcollaborativefiltering
AT zhenlongsun differentialprivacyframeworkforcollaborativefiltering
AT jianpeizhang differentialprivacyframeworkforcollaborativefiltering
_version_ 1725857327694741504

Similar Items