Privacy-preserving Bayesian network parameter learning

Authors:
Jianjie Ma;K. Sivakumar
Affiliations:
School of EECS, Washington State University, Pullman, WA;School of EECS, Washington State University, Pullman, WA
Venue:
CIMMACS'05 Proceedings of the 4th WSEAS international conference on Computational intelligence, man-machine systems and cybernetics
Year:
2005

Citing 7
Cited 1

On the design and quantification of privacy preserving data mining algorithms

PODS '01 Proceedings of the twentieth ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems
Tools for privacy preserving distributed data mining

ACM SIGKDD Explorations Newsletter
Limiting privacy breaches in privacy preserving data mining

Proceedings of the twenty-second ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems
k-anonymity: a model for protecting privacy

International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems
Collective Mining of Bayesian Networks from Distributed Heterogeneous Data

Knowledge and Information Systems
Privacy-preserving Bayesian network structure computation on distributed heterogeneous data

Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Privacy-Sensitive Bayesian Network Parameter Learning

ICDM '04 Proceedings of the Fourth IEEE International Conference on Data Mining

Secure multi party computation algorithm based on infinite product series

WSEAS Transactions on Information Science and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

Privacy is an important issue in data mining. Learning a Bayesian network (BN) from privacy sensitive data has been a recent research topic. In this paper, we propose to use a post randomization technique to learn Bayesian network parameters from distributed heterogeneous databases. The only required information from the data set is a set of sufficient statistics for learning Bayesian network parameters. The proposed method estimates the sufficient statistics from the randomized data. We show both theoretically and experimentally that, even with a large level of randomization, our method can learn the parameters accurately.