A hybrid multi-group privacy-preserving approach for building decision trees

Zhouxuan Teng; Wenliang Du

doi:10.1007/978-3-540-71701-0_30

A hybrid multi-group privacy-preserving approach for building decision trees

Zhouxuan Teng, Wenliang Du

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

12 Scopus citations

Abstract

In this paper, we study the privacy-preserving decision tree building problem on vertically partitioned data. We made two contributions. First, we propose a novel hybrid approach, which takes advantage of the strength of the two existing approaches, randomization and the secure multi-party computation (SMC), to balance the accuracy and efficiency constraints. Compared to these two existing approaches, our proposed approach can achieve much better accuracy than randomization approach and much reduced computation cost than SMC approach. We also propose a multi-group scheme that makes it flexible for data miners to control the balance between data mining accuracy and privacy. We partition attributes into groups, and develop a scheme to conduct group-based randomization to achieve better data mining accuracy. We have implemented and evaluated the proposed schemes for the ID3 decision tree algorithm.

Original language	English (US)
Title of host publication	Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings
Publisher	Springer Verlag
Pages	296-307
Number of pages	12
ISBN (Print)	9783540717003
DOIs	https://doi.org/10.1007/978-3-540-71701-0_30
State	Published - 2007
Event	11th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2007 - Nanjing, China Duration: May 22 2007 → May 25 2007

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4426 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	11th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2007
Country	China
City	Nanjing
Period	5/22/07 → 5/25/07

Keywords

Privacy
Randomization
SMC

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-540-71701-0_30

Fingerprint Dive into the research topics of 'A hybrid multi-group privacy-preserving approach for building decision trees'. Together they form a unique fingerprint.

Cite this

Teng, Z., & Du, W. (2007). A hybrid multi-group privacy-preserving approach for building decision trees. In Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings (pp. 296-307). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4426 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-540-71701-0_30

A hybrid multi-group privacy-preserving approach for building decision trees. / Teng, Zhouxuan; Du, Wenliang.

Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings. Springer Verlag, 2007. p. 296-307 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4426 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Teng, Z & Du, W 2007, A hybrid multi-group privacy-preserving approach for building decision trees. in Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4426 LNAI, Springer Verlag, pp. 296-307, 11th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2007, Nanjing, China, 5/22/07. https://doi.org/10.1007/978-3-540-71701-0_30

Teng Z, Du W. A hybrid multi-group privacy-preserving approach for building decision trees. In Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings. Springer Verlag. 2007. p. 296-307. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-71701-0_30

Teng, Zhouxuan ; Du, Wenliang. / A hybrid multi-group privacy-preserving approach for building decision trees. Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings. Springer Verlag, 2007. pp. 296-307 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{b308a5f83b5141538f91512f10d8e74c,

title = "A hybrid multi-group privacy-preserving approach for building decision trees",

abstract = "In this paper, we study the privacy-preserving decision tree building problem on vertically partitioned data. We made two contributions. First, we propose a novel hybrid approach, which takes advantage of the strength of the two existing approaches, randomization and the secure multi-party computation (SMC), to balance the accuracy and efficiency constraints. Compared to these two existing approaches, our proposed approach can achieve much better accuracy than randomization approach and much reduced computation cost than SMC approach. We also propose a multi-group scheme that makes it flexible for data miners to control the balance between data mining accuracy and privacy. We partition attributes into groups, and develop a scheme to conduct group-based randomization to achieve better data mining accuracy. We have implemented and evaluated the proposed schemes for the ID3 decision tree algorithm.",

keywords = "Privacy, Randomization, SMC",

author = "Zhouxuan Teng and Wenliang Du",

note = "Copyright: Copyright 2020 Elsevier B.V., All rights reserved.; 11th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2007 ; Conference date: 22-05-2007 Through 25-05-2007",

year = "2007",

doi = "10.1007/978-3-540-71701-0_30",

language = "English (US)",

isbn = "9783540717003",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "296--307",

booktitle = "Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings",

}

TY - GEN

T1 - A hybrid multi-group privacy-preserving approach for building decision trees

AU - Teng, Zhouxuan

AU - Du, Wenliang

PY - 2007

Y1 - 2007

N2 - In this paper, we study the privacy-preserving decision tree building problem on vertically partitioned data. We made two contributions. First, we propose a novel hybrid approach, which takes advantage of the strength of the two existing approaches, randomization and the secure multi-party computation (SMC), to balance the accuracy and efficiency constraints. Compared to these two existing approaches, our proposed approach can achieve much better accuracy than randomization approach and much reduced computation cost than SMC approach. We also propose a multi-group scheme that makes it flexible for data miners to control the balance between data mining accuracy and privacy. We partition attributes into groups, and develop a scheme to conduct group-based randomization to achieve better data mining accuracy. We have implemented and evaluated the proposed schemes for the ID3 decision tree algorithm.

AB - In this paper, we study the privacy-preserving decision tree building problem on vertically partitioned data. We made two contributions. First, we propose a novel hybrid approach, which takes advantage of the strength of the two existing approaches, randomization and the secure multi-party computation (SMC), to balance the accuracy and efficiency constraints. Compared to these two existing approaches, our proposed approach can achieve much better accuracy than randomization approach and much reduced computation cost than SMC approach. We also propose a multi-group scheme that makes it flexible for data miners to control the balance between data mining accuracy and privacy. We partition attributes into groups, and develop a scheme to conduct group-based randomization to achieve better data mining accuracy. We have implemented and evaluated the proposed schemes for the ID3 decision tree algorithm.

KW - Privacy

KW - Randomization

KW - SMC

UR - http://www.scopus.com/inward/record.url?scp=38049174606&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38049174606&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-71701-0_30

DO - 10.1007/978-3-540-71701-0_30

M3 - Conference contribution

AN - SCOPUS:38049174606

SN - 9783540717003

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 296

EP - 307

BT - Advances in Knowledge Discovery and Data Mining - 11th Pacific-Asia Conference, PAKDD 2007, Proceedings

PB - Springer Verlag

T2 - 11th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2007

Y2 - 22 May 2007 through 25 May 2007

ER -