Optimizing Batch Linear Queries under Exact and Approximate Differential Privacy

被引:22
|
作者
Yuan, Ganzhao [1 ]
Zhang, Zhenjie [2 ]
Winslett, Marianne [2 ,3 ]
Xiao, Xiaokui [4 ]
Yang, Yin [5 ]
Hao, Zhifeng [6 ,7 ]
机构
[1] S China Univ Technol, Dept Math, Guangzhou, Guangdong, Peoples R China
[2] Adv Digital Sci Ctr, Singapore, Singapore
[3] Univ Illinois, Dept Comp Sci, Urbana, IL 61801 USA
[4] Nanyang Technol Univ, Sch Comp Engn, Singapore 639798, Singapore
[5] Hamad Bin Khalifa Univ, Div Informat & Commun Technol, Ar Rayyan, Qatar
[6] S China Univ Technol, Sch Comp Sci & Engn, Guangzhou, Guangdong, Peoples R China
[7] Guangdong Univ Technol, Fac Comp Sci, Guangzhou, Guangdong, Peoples R China
来源
ACM TRANSACTIONS ON DATABASE SYSTEMS | 2015年 / 40卷 / 02期
关键词
Theory; Algorithms; Experimentation; Linear counting query; differential privacy; low rank; matrix approximation; augmented Lagrangian multiplier algorithm; ALGORITHM; MECHANISM; OPTIMIZATION; VOLUME; NOISE;
D O I
10.1145/2699501
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Differential privacy is a promising privacy-preserving paradigm for statistical query processing over sensitive data. It works by injecting random noise into each query result such that it is provably hard for the adversary to infer the presence or absence of any individual record from the published noisy results. The main objective in differentially private query processing is to maximize the accuracy of the query results while satisfying the privacy guarantees. Previous work, notably Li et al. [2010], has suggested that, with an appropriate strategy, processing a batch of correlated queries as a whole achieves considerably higher accuracy than answering them individually. However, to our knowledge there is currently no practical solution to find such a strategy for an arbitrary query batch; existing methods either return strategies of poor quality (often worse than naive methods) or require prohibitively expensive computations for even moderately large domains. Motivated by this, we propose a low-rank mechanism (LRM), the first practical differentially private technique for answering batch linear queries with high accuracy. LRM works for both exact (i.e., epsilon-) and approximate (i.e., (epsilon, delta)-) differential privacy definitions. We derive the utility guarantees of LRM and provide guidance on how to set the privacy parameters, given the user's utility expectation. Extensive experiments using real data demonstrate that our proposed method consistently outperforms state-of-the-art query processing solutions under differential privacy, by large margins.
引用
收藏
页数:47
相关论文
共 50 条
  • [1] Optimizing Linear Counting Queries Under Differential Privacy
    Li, Chao
    Hay, Michael
    Rastogi, Vibhor
    Miklau, Gerome
    McGregor, Andrew
    PODS 2010: PROCEEDINGS OF THE TWENTY-NINTH ACM SIGMOD-SIGACT-SIGART SYMPOSIUM ON PRINCIPLES OF DATABASE SYSTEMS, 2010, : 123 - 134
  • [2] Low-Rank Mechanism: Optimizing Batch Queries under Differential Privacy
    Yuan, Ganzhao
    Zhang, Zhenjie
    Winslett, Marianne
    Xiao, Xiaokui
    Yang, Yin
    Hao, Zhifeng
    PROCEEDINGS OF THE VLDB ENDOWMENT, 2012, 5 (11): : 1352 - 1363
  • [3] The matrix mechanism: optimizing linear counting queries under differential privacy
    Chao Li
    Gerome Miklau
    Michael Hay
    Andrew McGregor
    Vibhor Rastogi
    The VLDB Journal, 2015, 24 : 757 - 781
  • [4] The matrix mechanism: optimizing linear counting queries under differential privacy
    Li, Chao
    Miklau, Gerome
    Hay, Michael
    McGregor, Andrew
    Rastogi, Vibhor
    VLDB JOURNAL, 2015, 24 (06): : 757 - 781
  • [5] Optimizing error of high-dimensional statistical queries under differential privacy
    McKenna, Ryan
    Miklau, Gerome
    Hay, Michael
    Machanavajjhala, Ashwin
    PROCEEDINGS OF THE VLDB ENDOWMENT, 2018, 11 (10): : 1206 - 1219
  • [6] Orthogonal Mechanism for Answering Batch Queries with Differential Privacy
    Huang, Dong
    Han, Shuguo
    Li, Xiaoli
    Yu, Philip S.
    PROCEEDINGS OF THE 27TH INTERNATIONAL CONFERENCE ON SCIENTIFIC AND STATISTICAL DATABASE MANAGEMENT, 2015,
  • [7] Linear Queries Estimation with Local Differential Privacy
    Bassily, Raef
    22ND INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 89, 2019, 89 : 721 - 729
  • [8] A workload-adaptive mechanism for linear queries under local differential privacy
    McKenna, Ryan
    Maity, Raj Kumar
    Mazumdar, Arya
    Miklau, Gerome
    PROCEEDINGS OF THE VLDB ENDOWMENT, 2020, 13 (11): : 1905 - 1918
  • [9] Convex Optimization for Linear Query Processing under Approximate Differential Privacy
    Yuan, Ganzhao
    Yang, Yin
    Zhang, Zhenjie
    Hao, Zhifeng
    KDD'16: PROCEEDINGS OF THE 22ND ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2016, : 2005 - 2014
  • [10] Achieving Accuracy Guarantee for Answering Batch Queries with Differential Privacy
    Huang, Dong
    Han, Shuguo
    Li, Xiaoli
    ADVANCES IN KNOWLEDGE DISCOVERY AND DATA MINING, PART II, 2015, 9078 : 305 - 316