Accurate classification models for distributed mining of privately preserved data

M. Sumana, K. S. Hareesha

Research output: Contribution to journalReview article

Abstract

Data maintained at various sectors, needs to be mined to derive useful inferences. Larger part of the data is sensitive and not to be revealed while mining. Current methods perform privacy preservation classification either by randomizing, perturbing or anonymizing the data during mining. These forms of privacy preserving mining work well for data centralized at a single site. Moreover the amount of information hidden during mining is not sufficient. When perturbation approaches are used, data reconstruction is a major challenge. This paper aims at modeling classifiers for data distributed across various sites with respect to the same instances. The homomorphic and probabilistic property of Paillier is used to perform secure product, mean and variance calculations. The secure computations are performed without any intermediate data or the sensitive data at multiple sites being revealed. It is observed that the accuracy of the classifiers modeled is almost equivalent to the non-privacy preserving classifiers. Secure protocols require reduced computation time and communication cost.

Original languageEnglish
Pages (from-to)58-73
Number of pages16
JournalInternational Journal of Information Security and Privacy
Volume10
Issue number4
DOIs
Publication statusPublished - 01-10-2016

Fingerprint

Classifiers
Data mining
Communication
Costs

All Science Journal Classification (ASJC) codes

  • Information Systems

Cite this

@article{151f700ba04d48129c7671858d94f165,
title = "Accurate classification models for distributed mining of privately preserved data",
abstract = "Data maintained at various sectors, needs to be mined to derive useful inferences. Larger part of the data is sensitive and not to be revealed while mining. Current methods perform privacy preservation classification either by randomizing, perturbing or anonymizing the data during mining. These forms of privacy preserving mining work well for data centralized at a single site. Moreover the amount of information hidden during mining is not sufficient. When perturbation approaches are used, data reconstruction is a major challenge. This paper aims at modeling classifiers for data distributed across various sites with respect to the same instances. The homomorphic and probabilistic property of Paillier is used to perform secure product, mean and variance calculations. The secure computations are performed without any intermediate data or the sensitive data at multiple sites being revealed. It is observed that the accuracy of the classifiers modeled is almost equivalent to the non-privacy preserving classifiers. Secure protocols require reduced computation time and communication cost.",
author = "M. Sumana and Hareesha, {K. S.}",
year = "2016",
month = "10",
day = "1",
doi = "10.4018/IJISP.2016100104",
language = "English",
volume = "10",
pages = "58--73",
journal = "International Journal of Information Security and Privacy",
issn = "1930-1650",
publisher = "IGI Global Publishing",
number = "4",

}

Accurate classification models for distributed mining of privately preserved data. / Sumana, M.; Hareesha, K. S.

In: International Journal of Information Security and Privacy, Vol. 10, No. 4, 01.10.2016, p. 58-73.

Research output: Contribution to journalReview article

TY - JOUR

T1 - Accurate classification models for distributed mining of privately preserved data

AU - Sumana, M.

AU - Hareesha, K. S.

PY - 2016/10/1

Y1 - 2016/10/1

N2 - Data maintained at various sectors, needs to be mined to derive useful inferences. Larger part of the data is sensitive and not to be revealed while mining. Current methods perform privacy preservation classification either by randomizing, perturbing or anonymizing the data during mining. These forms of privacy preserving mining work well for data centralized at a single site. Moreover the amount of information hidden during mining is not sufficient. When perturbation approaches are used, data reconstruction is a major challenge. This paper aims at modeling classifiers for data distributed across various sites with respect to the same instances. The homomorphic and probabilistic property of Paillier is used to perform secure product, mean and variance calculations. The secure computations are performed without any intermediate data or the sensitive data at multiple sites being revealed. It is observed that the accuracy of the classifiers modeled is almost equivalent to the non-privacy preserving classifiers. Secure protocols require reduced computation time and communication cost.

AB - Data maintained at various sectors, needs to be mined to derive useful inferences. Larger part of the data is sensitive and not to be revealed while mining. Current methods perform privacy preservation classification either by randomizing, perturbing or anonymizing the data during mining. These forms of privacy preserving mining work well for data centralized at a single site. Moreover the amount of information hidden during mining is not sufficient. When perturbation approaches are used, data reconstruction is a major challenge. This paper aims at modeling classifiers for data distributed across various sites with respect to the same instances. The homomorphic and probabilistic property of Paillier is used to perform secure product, mean and variance calculations. The secure computations are performed without any intermediate data or the sensitive data at multiple sites being revealed. It is observed that the accuracy of the classifiers modeled is almost equivalent to the non-privacy preserving classifiers. Secure protocols require reduced computation time and communication cost.

UR - http://www.scopus.com/inward/record.url?scp=84987819166&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84987819166&partnerID=8YFLogxK

U2 - 10.4018/IJISP.2016100104

DO - 10.4018/IJISP.2016100104

M3 - Review article

AN - SCOPUS:84987819166

VL - 10

SP - 58

EP - 73

JO - International Journal of Information Security and Privacy

JF - International Journal of Information Security and Privacy

SN - 1930-1650

IS - 4

ER -