Differentially private high-dimensional data publication via grouping and truncating techniques

Abstract

The count of one column for high-dimensional datasets, i.e., the number of records containing this column, has been widely used in numerous applications such as analyzing popular spots based on check-in location information and mining valuable items from shopping records. However, this poses a privacy threat when directly publishing this information. Differential privacy (DP), as a notable paradigm for strong privacy guarantees, is thereby adopted to publish all column counts. Prior studies have verified that truncating records or grouping columns can effectively improve the accuracy of published results. To leverage the advantages of the two techniques, we combine these studies to further boost the accuracy of published results. However, the traditional penalty function, which measures the error imported by a given pair of parameters including truncating length and group size, is so sensitive that the derived parameters deviate from the optimal parameters significantly. To output preferable parameters, we first design a smart penalty function that is less sensitive than the traditional function. Moreover, a two-phase selection method is proposed to compute these parameters efficiently, together with the improvement in accuracy. Extensive experiments on a broad spectrum of real-world datasets validate the effectiveness of our proposals.

Keywords

differential privacy / high-dimensional data / truncation optimization / grouping / penalty function

References

[1] Dwork C, McSherry F, Nissim K, Smith A. Calibrating noise to sensitivity in private data analysis. In: Proceedings of Theory of Cryptography Conference. 2006, 265–284

[2] Dwork C, Rothblum G N, Vadhan S P. Boosting and differential privacy. In: Proceedings of Annual IEEE Symposium on Foundations of Computer Science. 2010, 51–60

[3] McSherry F, Talwar K. Mechanism design via differential privacy. In: Proceedings of Annual IEEE Symposium on Foundations of Computer Science. 2007, 94–103

[4] McSherry F. Privacy integrated queries: an extensible platform for privacy-preserving data analysis. In: Proceedings of ACM SIGMOD International Conference on Management of Data. 2009, 19–30

[5] Rastogi V, Nath S. Differentially private aggregation of distributed time-series with transformation and encryption. In: Proceedings of ACM SIGMOD International Conference on Management of Data. 2010, 735–746

[6] Kellaris G, Papadopoulos S. Practical differential privacy via grouping and smoothing. In: Proceedings of International Conference on Very Large Data Bases. 2013, 301–312

[7] Day W Y, Li N H. Differentially private publishing of high-dimensional data using sensitivity control. In: Proceedings of ACM Symposium on Information, Computer and Communications Security. 2015, 451–462

[8] Day W Y, Li N H, Lyu M. Publishing graph degree distribution with node differential privacy. In: Proceedings of ACM SIGMOD International Conference on Management of Data. 2016, 123–138

[9] Hardt M, Ligett K, McSherry F. A simple and practical algorithm for differentially private data release. In: Proceedings of Annual Conference on Neural Information Processing Systems. 2012, 2348–2356

[10] Xiao X K, Wang G Z, Gehrke J. Differential privacy via wavelet transforms. In: Proceedings of International Conference on Data Engineering. 2010, 225–236

[11] Zhang X J, Chen R, Xu J L, Meng X F, Xie Y T. Towards accurate histogram publication under differential privacy. In: Proceedings of SIAM International Conference on Data Mining. 2014, 587–595

[12] Xu J, Zhang Z J, Xiao X K, Yang Y, Yu G. Differentially private histogram publication. In: Proceedings of IEEE International Conference on Data Engineering. 2012, 32–43

[13] Hay M, Rastogi V, Miklau G, Suciu D. Boosting the accuracy of differentially private histograms through consistency. In: Proceedings of International Conference on Very Large Data Bases. 2010, 1021–1032

[14] Li C, Hay M, Rastogi V, Miklau G, McGregor A. Optimizing linear counting queries under differential privacy. In: Proceedings of ACM SIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems. 2010, 123–134

[15] Qardaji W H, Yang W N, Li N H. Understanding hierarchical methods for differentially private histograms. In: Proceedings of International Conference on Very Large Data Bases. 2013, 1954–1965

[16] Li C, Hay M, Miklau G, Wang Y. A data- and workload-aware query answering algorithm for range queries under differential privacy. In: Proceedings of International Conference on Very Large Data Bases. 2014, 341–352

[17] Chen R, Mohammed N, Fung C M, Desai B C, Xiong L. Publishing set-valued data via differential privacy. In: Proceedings of International Conference on Very Large Data Bases. 2011, 1087–1098

[18] Chen R, Ács G, Castelluccia C. Differentially private sequential data publication via variable-length n-grams. In: Proceedings of ACM Conference on Computer and Communications Security. 2012, 638–649

[19] Zhang J, Cormode G, Procopiuc C M, Srivastava D, Xiao X K. Privbayes: private data release via bayesian networks. In: Proceedings of ACM SIGMOD International Conference on Management of Data. 2014, 1423–1434

[20] Qardaji W H, Yang W N, Li N H. Priview: practical differentially private release of marginal contingency tables. In: Proceedings of ACM SIGMOD International Conference on Management of Data. 2014, 1435–1446

[21] He X, Cormode G, Machanavajjhala A, Procopiuc C M, Srivastava D. DPT: differentially private trajectory synthesis using hierarchical reference systems. In: Proceedings of International Conference on Very Large Data Bases. 2015, 1154–1165

[22] Kasiviswanathan S P, Nissim K, Raskhodnikova S, Smith A. Analyzing graphs with node differential privacy. In: Proceedings of Theory of Cryptography Conference. 2013, 457–476

RIGHTS & PERMISSIONS

2018 Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature