Third Party Privacy Preserving Protocol for Perturbation Based Classification of Vertically Fragmented Data Bases

TL;DR

This paper introduces a privacy-preserving protocol for classifying vertically fragmented data using perturbed data and a third-party model, achieving high accuracy without revealing original data.

Contribution

It combines data perturbation with third-party protocols to enable accurate Naive Bayes classification on decentralized, disguised datasets, enhancing security in distributed data mining.

Findings

High classification accuracy with perturbed data

Effective privacy preservation in decentralized databases

Compatibility with existing classification techniques

Abstract

Privacy is become major issue in distributed data mining. In the literature we can found many proposals of privacy preserving which can be divided into two major categories that is trusted third party and multiparty based privacy protocols. In case of trusted third party models the conventional asymmetric cryptographic based techniques will be used and in case of multi party based protocols data perturbed to make sure no other party to understand original data. In order to enhance security features by combining strengths of both models in this paper, we propose to use data perturbed techniques in third party privacy preserving protocol to conduct the classification on vertically fragmented data bases. Specially, we present a method to build Naive Bayes classification from the disguised and decentralized databases. In order to perform classification we propose third party protocol for secure computations. We conduct experiments to compare the accuracy of our Naive Bayes with the one built from the original undisguised data. Our results show that although the data are disguised and decentralized, our method can still achieve fairly high accuracy.