WaveCluster with differential privacy

Ling Chen, Ting Yu, Rada Chirkova

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

7 Citations (Scopus)

Abstract

WaveCluster is an important family of grid-based clustering algorithms that are capable of finding clusters of arbitrary shapes. In this paper, we investigate techniques to perform WaveCluster while ensuring differential privacy. Our goal is to develop a general technique for achieving differential privacy on WaveCluster that accommodates different wavelet transforms. We show that straightforward techniques based on synthetic data generation and introduction of random noise when quantizing the data, though generally preserving the distribution of data, often introduce too much noise to preserve useful clusters. We then propose two optimized techniques, PrivTHR and PrivTHREM, which can significantly reduce data distortion during two key steps of WaveCluster: the quantization step and the significant grid identification step. We conduct extensive experiments based on four datasets that are particularly interesting in the context of clustering, and show that PrivTHR and PrivTHREM achieve high utility when privacy budgets are properly allocated, conforming to our theoretical analysis.

Original languageEnglish
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
PublisherAssociation for Computing Machinery
Pages1011-1020
Number of pages10
Volume19-23-Oct-2015
ISBN (Print)9781450337946
DOIs
Publication statusPublished - 17 Oct 2015
Event24th ACM International Conference on Information and Knowledge Management, CIKM 2015 - Melbourne, Australia
Duration: 19 Oct 201523 Oct 2015

Other

Other24th ACM International Conference on Information and Knowledge Management, CIKM 2015
CountryAustralia
CityMelbourne
Period19/10/1523/10/15

Fingerprint

Privacy
Grid
Theoretical analysis
Clustering
Clustering algorithm
Quantization
Experiment
Wavelet transform

Keywords

  • Clustering
  • Differential privacy
  • Wavelet transform

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Decision Sciences(all)

Cite this

Chen, L., Yu, T., & Chirkova, R. (2015). WaveCluster with differential privacy. In International Conference on Information and Knowledge Management, Proceedings (Vol. 19-23-Oct-2015, pp. 1011-1020). Association for Computing Machinery. https://doi.org/10.1145/2806416.2806546

WaveCluster with differential privacy. / Chen, Ling; Yu, Ting; Chirkova, Rada.

International Conference on Information and Knowledge Management, Proceedings. Vol. 19-23-Oct-2015 Association for Computing Machinery, 2015. p. 1011-1020.

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

Chen, L, Yu, T & Chirkova, R 2015, WaveCluster with differential privacy. in International Conference on Information and Knowledge Management, Proceedings. vol. 19-23-Oct-2015, Association for Computing Machinery, pp. 1011-1020, 24th ACM International Conference on Information and Knowledge Management, CIKM 2015, Melbourne, Australia, 19/10/15. https://doi.org/10.1145/2806416.2806546
Chen L, Yu T, Chirkova R. WaveCluster with differential privacy. In International Conference on Information and Knowledge Management, Proceedings. Vol. 19-23-Oct-2015. Association for Computing Machinery. 2015. p. 1011-1020 https://doi.org/10.1145/2806416.2806546
Chen, Ling ; Yu, Ting ; Chirkova, Rada. / WaveCluster with differential privacy. International Conference on Information and Knowledge Management, Proceedings. Vol. 19-23-Oct-2015 Association for Computing Machinery, 2015. pp. 1011-1020
@inproceedings{5c77debf30de45c5b2fc1f09e571d293,
title = "WaveCluster with differential privacy",
abstract = "WaveCluster is an important family of grid-based clustering algorithms that are capable of finding clusters of arbitrary shapes. In this paper, we investigate techniques to perform WaveCluster while ensuring differential privacy. Our goal is to develop a general technique for achieving differential privacy on WaveCluster that accommodates different wavelet transforms. We show that straightforward techniques based on synthetic data generation and introduction of random noise when quantizing the data, though generally preserving the distribution of data, often introduce too much noise to preserve useful clusters. We then propose two optimized techniques, PrivTHR and PrivTHREM, which can significantly reduce data distortion during two key steps of WaveCluster: the quantization step and the significant grid identification step. We conduct extensive experiments based on four datasets that are particularly interesting in the context of clustering, and show that PrivTHR and PrivTHREM achieve high utility when privacy budgets are properly allocated, conforming to our theoretical analysis.",
keywords = "Clustering, Differential privacy, Wavelet transform",
author = "Ling Chen and Ting Yu and Rada Chirkova",
year = "2015",
month = "10",
day = "17",
doi = "10.1145/2806416.2806546",
language = "English",
isbn = "9781450337946",
volume = "19-23-Oct-2015",
pages = "1011--1020",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - WaveCluster with differential privacy

AU - Chen, Ling

AU - Yu, Ting

AU - Chirkova, Rada

PY - 2015/10/17

Y1 - 2015/10/17

N2 - WaveCluster is an important family of grid-based clustering algorithms that are capable of finding clusters of arbitrary shapes. In this paper, we investigate techniques to perform WaveCluster while ensuring differential privacy. Our goal is to develop a general technique for achieving differential privacy on WaveCluster that accommodates different wavelet transforms. We show that straightforward techniques based on synthetic data generation and introduction of random noise when quantizing the data, though generally preserving the distribution of data, often introduce too much noise to preserve useful clusters. We then propose two optimized techniques, PrivTHR and PrivTHREM, which can significantly reduce data distortion during two key steps of WaveCluster: the quantization step and the significant grid identification step. We conduct extensive experiments based on four datasets that are particularly interesting in the context of clustering, and show that PrivTHR and PrivTHREM achieve high utility when privacy budgets are properly allocated, conforming to our theoretical analysis.

AB - WaveCluster is an important family of grid-based clustering algorithms that are capable of finding clusters of arbitrary shapes. In this paper, we investigate techniques to perform WaveCluster while ensuring differential privacy. Our goal is to develop a general technique for achieving differential privacy on WaveCluster that accommodates different wavelet transforms. We show that straightforward techniques based on synthetic data generation and introduction of random noise when quantizing the data, though generally preserving the distribution of data, often introduce too much noise to preserve useful clusters. We then propose two optimized techniques, PrivTHR and PrivTHREM, which can significantly reduce data distortion during two key steps of WaveCluster: the quantization step and the significant grid identification step. We conduct extensive experiments based on four datasets that are particularly interesting in the context of clustering, and show that PrivTHR and PrivTHREM achieve high utility when privacy budgets are properly allocated, conforming to our theoretical analysis.

KW - Clustering

KW - Differential privacy

KW - Wavelet transform

UR - http://www.scopus.com/inward/record.url?scp=84958248973&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958248973&partnerID=8YFLogxK

U2 - 10.1145/2806416.2806546

DO - 10.1145/2806416.2806546

M3 - Conference contribution

AN - SCOPUS:84958248973

SN - 9781450337946

VL - 19-23-Oct-2015

SP - 1011

EP - 1020

BT - International Conference on Information and Knowledge Management, Proceedings

PB - Association for Computing Machinery

ER -