一种基于差分隐私和时序的推荐系统模型研究

doi:10.3969/j.issn.0372-2112.2017.09.001

电子学报 ›› 2017, Vol. 45 ›› Issue (9): 2057-2064.DOI: 10.3969/j.issn.0372-2112.2017.09.001

• 学术论文 • 下一篇

一种基于差分隐私和时序的推荐系统模型研究

范利云^1,2, 左万利^1,2, 王英^1,2, 王鑫³

1. 吉林大学计算机科学与技术学院, 吉林长春 130012;
2. 吉林大学符号计算与知识工程教育部重点实验室, 吉林长春 130012;
3. 长春工程学院计算机技术与工程学院, 吉林长春 130012

收稿日期:2016-04-11 修回日期:2017-02-14 出版日期:2017-09-25
- 通讯作者:
- 左万利
- 作者简介:
- 范利云,女,1990年出生,河南安阳人,现为吉林大学计算机科学与技术学院硕士研究生,从事机器学习、数据挖掘等有关研究.E-mail:fanlyjlu@163.com
- 基金资助:
- 国家自然科学基金 (No.60973040，No.61300148，No.61602057）; 吉林省科技发展计划 (No.20130206051GX，No.20130522112JH，No.20170520059JH）

Research on Recommender System Model Based on Differential Privacy and Time Series

FAN Li-yun^1,2, ZUO Wan-li^1,2, WANG Ying^1,2, WANG Xin³

1. College of Computer Science and Technology, Jilin University, Changchun, Jilin 130012, China;
2. Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Jilin University, Changchun, Jilin 130012, China;
3. School of Computer Technology and Engineering, Changchun Institute of Technology, Changchun, Jilin 130012, China

Received:2016-04-11 Revised:2017-02-14 Online:2017-09-25 Published:2017-09-25

摘要/Abstract

摘要： 推荐系统的建立依赖用户的个人隐私信息，攻击者可以通过推荐的结果对用户的状态和行为进行预测.目前，虽然有对基于协同过滤近邻隐私保护的研究，但是对基于模型的隐私保护的关注度并不够高.差分隐私理论定义了一个相当严格的防攻击模型，通过添加噪声使数据失真达到隐私保护的目的，而且用户的兴趣存在兴趣漂移问题，对推荐效果造成影响，因此，提出基于差分隐私理论和时序理论构建基于模型的推荐系统.首先，根据差分隐私理论，给用户的评分数据增加小波动的符合Laplace分布的噪声，增大待分解矩阵的安全系数；然后，在随机梯度下降模型的基础上，将时序因子建模为时间权重，提高模型的准确性.实验证明该算法的准确性，并且为增强隐私研究提供了新的思路.

Abstract: Recommender system is established on users' private information.However,based on results of recommender system,attackers can predict users' states and behaviors.At present,although some researchers focus on collaborative filtering neighbor theory to preserve users' privacy,very few researchers pay enough attention to the model-based privacy-preserving.Differential privacy offers a strong degree of privacy protection by adding noise.And there is interest drift in users' interest.So this paper proposes a recommender system model based on differential privacy theory and time series theory.Firstly,according to differential privacy theory,we add some Laplace-distribution-fitted noises into users' score data to enlarge safety factor in factorization matrix.Then based on Stochastic gradient descent model,we model time series factor as time weight to improve the accuracy of the model.Experimental results demonstrate the accuracy of the algorithm,which provides a valuable perspective for privacy-preserving recommender research.

Key words: recommender system, non-negative matrix factorization, Stochastic gradient descent, differential privacy, time series

中图分类号:

TP18

范利云, 左万利, 王英, 王鑫. 一种基于差分隐私和时序的推荐系统模型研究[J]. 电子学报, 2017, 45(9): 2057-2064.

FAN Li-yun, ZUO Wan-li, WANG Ying, WANG Xin. Research on Recommender System Model Based on Differential Privacy and Time Series[J]. Acta Electronica Sinica, 2017, 45(9): 2057-2064.

参考文献

[1] Weinsberg U,Bhagat S,Ioannidis S,et al.Blurme:inferring and obfuscating user gender based on ratings[A].Proceedings of the Sixth ACM Conference on Recommender Systems[C].Dublin,Ireland:ACM,2012.195-202.
[2] Calandrino J A,Kilzer A,Narayanan A,et al."You might also like:" privacy risks of collaborative filtering[A].Proceedings of the 2011 IEEE Symposium on Security and Privacy[C].Washington,DC,USA:IEEE,2011.231-246.
[3] 熊平,朱天清,王晓峰.差分隐私保护及其应用[J].计算机学报,2014,37(1):101-122. Xiong P,Zhu T Q,Wang X F.A survey on differential privacy and applications[J].Chinese Journal of Computers,2014,37(1):101-122.(in Chinese)
[4] Mcsherry F,Mironov I.Differentially private recommender systems:building privacy into the net[A].Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining[C].Paris,France:ACM,2009.627-636.
[5] Dwork C.Differential privacy[A].Proceedings of the 33rd International Conference on Automata,Languages and Programming[C].Venice,Italy:Springer,2006.1-12.
[6] Resnick P,Varian H R.Recommender systems[J].Communications of the ACM,1997,40(3):56-58.
[7] Bogers T,Koolen M.Second workshop on new trends in content-based recommender systems (CBRecSys 2015)[A].Proceedings of the 9th ACM Conference on Recommender Systems[C].Vienna,Austria:ACM,2015.339-340.
[8] Bogers T,Koolen M.Report on RecSys 2015 workshop on new trends in content-based recommender systems[J].ACM SIGIR Forum,2016,49(2):141-146.
[9] Anava O,Golan S,Golbandi N,et al.Budget-constrained item cold-start handling in collaborative filtering recommenders via optimal design[A].Proceedings of the 24th International Conference on World Wide Web[C].Florence,Italy:ACM,2015.45-54.
[10] Jiang X,Liu W,Cao L,et al.Coupled collaborative filtering for context-aware recommendation[A].Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence[C].Texas,USA:AAAI,2015.4172-4173.
[11] Liu Z,Wang Y X,Smola A.Fast differentially private matrix factorization[A].Proceedings of the 9th ACM Conference on Recommender Systems[C].Vienna,Austria:ACM,2015.171-178.
[12] 印桂生,张亚楠,董宇欣,等.基于受限信任关系和概率分解矩阵的推荐[J].电子学报,2013,42(5):904-911. Yin G S,Zhang Y N,Dong Y X,et al.A constrained trust recommendation using probabilistic matrix factorization[J].Acta Electronica Sinica,2013,42(5):904-911.(in Chinese)
[13] Zhang W,Wang J.A collective bayesian poisson factorization model for cold-start local event recommendation[A].Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining[C].Sydney,Australia:ACM,2015.1455-1464..
[14] Kouki P,Fakhraei S,Foulds J,et al.HyPER:A flexible and extensible probabilistic framework for hybrid recommender systems[A].Proceedings of the 9th ACM Conference on Recommender Systems[C].Vienna,Austria:ACM,2015.99-106.
[15] Boutet A,Frey D,Guerraoui R,et al.Privacy-preserving distributed collaborative filtering[A].Proceedings of the Second International Conference,NETYS 2014[C].Marrakech,Morocco:Springer,2014.169-184.
[16] Vallet D,Friedman A,Berkovsky S.Matrix factorization without user data retention[A].Proceedings of the 18th Pacific-Asia Conference on Knowledge Discovery and Data Mining[C].Tainan,Taiwan:Springer,2014.569-580.
[17] Nikolaenko V,Ioannidis S,Weinsberg U,et al.Privacy-preserving matrix factorization[A].Proceedings of the 2013 ACM SIGSAC Conference on Computer & Communications Security[C].Berlin,Germany:ACM,2013.801-812.
[18] Berkovsky S,Kuflik T,Ricci F.The impact of data obfuscation on the accuracy of collaborative filtering[J].Expert Systems with Applications,2012,39(5):5033-5042.
[19] Zhu X,Sun Y.Differential privacy for collaborative filtering recommender algorithm[A].Proceedings of the 2016 ACM on International Workshop on Security And Privacy Analytics[C].New Orleans,Louisiana,USA:ACM,2016.9-16.
[20] Koren Y.Collaborative filtering with temporal dynamics.Commun ACM[J].Communications of the ACM,2010,53(4):89-97.
[21] Xiong L,Chen X,Huang T,et al.Temporal collaborative filtering with bayesian probabilistic tensor factorization[A].Proceedings of the 2010 SIAM International Conference on Data Mining[C].Columbus,Ohio,USA:DBLP,2010.211-222.
[22] Khoshneshin M,Street W N.Incremental collaborative filtering via evolutionary co-clustering[A].Proceedings of the Fourth ACM Conference on Recommender systems[C].Barcelona,Spain:ACM,2010.325-328.
[23] Dwork C,Mcsherry F,Nissim K.Calibrating noise to sensitivity in private data analysis[A].Proceedings of the Third Conference on Theory of Cryptography[C].New York,USA:Springer,2006.265-284.
[24] Dwork C.Differential privacy:A survey of results[A].Proceedings of the 5th International Conference on Theory and Applications of Models of Computation[C].Xian,China:Springer,2008.1-19.
[25] Berlioz A,Friedman A,Kaafar M A,et al.Applying differential privacy to matrix factorization[A].Proceedings of the 9th ACM Conference on Recommender Systems[C].Vienna,Austria:ACM,2015.107-114.
[26] Rennie J D M,Srebro N.Fast maximum margin matrix factorization for collaborative prediction[A].Proceedings of the 22nd International Conference on Machine Learning[C].Bonn,Germany:ACM,2005.713-719.
[27] Salakhutdinov R,Mnih A.Probabilistic matrix factorization[A].Proceedings of Advances in Neural Information Processing Systems[C].Vancouver,Canada:NIPS,2007.1257-1264.

一种基于差分隐私和时序的推荐系统模型研究

Research on Recommender System Model Based on Differential Privacy and Time Series

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	康海燕, 冀源蕊. 基于本地化差分隐私的时序位置发布方案研究[J]. 电子学报, 2022, 50(9): 2222-2232.
[2]	张少波, 原刘杰, 毛新军, 朱更明. 基于本地差分隐私的K-modes聚类数据隐私保护方法[J]. 电子学报, 2022, 50(9): 2181-2188.
[3]	张帅, 高旻, 文俊浩, 熊庆宇, 唐旭. 基于自监督学习的去流行度偏差推荐方法[J]. 电子学报, 2022, 50(10): 2361-2371.
[4]	袁水莲, 皮德常, 胥萌. 基于差分隐私的轨迹隐私保护方法[J]. 电子学报, 2021, 49(7): 1266-1273.
[5]	郑孝遥, 罗永龙, 汪祥舜, 孙丽萍, 陈付龙, 胡桂银, 汪小寒. 基于位置服务的分布式差分隐私推荐方法研究[J]. 电子学报, 2021, 49(1): 99-110.
[6]	李东瑾, 杨瑞娟, 李晓柏, 董睿杰. 基于核协同表示与鉴别投影的辐射源调制识别[J]. 电子学报, 2020, 48(9): 1695-1702.
[7]	郭景峰, 朱晓松, 李爽. 直播电视推荐系统的评分预测算法研究[J]. 电子学报, 2020, 48(9): 1735-1740.
[8]	李东瑾, 杨瑞娟, 李晓柏, 董睿杰. 基于栈式稀疏降噪自编码网络的辐射源调制识别[J]. 电子学报, 2020, 48(6): 1198-1204.
[9]	黄名选, 蒋曹清. 基于项权值排序挖掘的跨语言查询扩展[J]. 电子学报, 2020, 48(3): 568-576.
[10]	陈思, 付安民, 柯海峰, 苏铓, 孙怀江. MCDP:基于神经网络的多集群分布式差分隐私数据发布方法[J]. 电子学报, 2020, 48(12): 2297-2303.
[11]	方晨, 郭渊博, 王娜, 甄帅辉, 唐国栋. 基于生成对抗网络的差分隐私数据发布方法[J]. 电子学报, 2020, 48(10): 1983-1992.
[12]	彭慧丽, 金凯忠, 付聪聪, 付楠, 张啸剑. 基于序列格的隐私时序模式挖掘方法[J]. 电子学报, 2020, 48(1): 153-163.
[13]	刘露, 胡封晔, 牛亮, 彭涛. 异质网络中基于节点影响力的相似度度量方法[J]. 电子学报, 2019, 47(9): 1929-1936.
[14]	张二华, 王明合, 唐振民. 加性噪声条件下鲁棒说话人确认[J]. 电子学报, 2019, 47(6): 1244-1250.
[15]	贺超波, 汤庸, 张琼, 刘双印, 刘海. 基于增量式鲁棒非负矩阵分解的短文本在线聚类[J]. 电子学报, 2019, 47(5): 1086-1093.