Key Compromise Resilient Privacy Provisioning in Vertically Partitioned Data

!!!! Bi-Annual Double Blind Peer Reviewed Refereed Journal !!!!

!!!! Open Access Journal !!!!

Category: 
Technology
Author: 

S KumaraSwamy - Dept. of CSE, University Visvesvaraya College of Engineering, Bangalore University

Manjula S H - Dept. of CSE, University Visvesvaraya College of Engineering, Bangalore University

K R Venugopal - Dept. of CSE, University Visvesvaraya College of Engineering, Bangalore University

Iyengar S S - Professor, Dept. of CSE, Florida International University, USA

L M Patnaik - Honorary Professor, Dept. of CSE, Indian Institute of Science, Bangalore

Abstract: 

The mounting concerns related to privacy preservation of confidential data, organizations enforce the use of cryptic techniques to preserve the privacy of the data to be used for mining. In this paper a secure multi party communication protocol is described eliminating overheads and vulnerabilities introduced due to key distribution. The protocol described enhances the accuracy of mining analysis a desired feature of any Privacy Preserving Data Mining (PPDM) system by adopting the C5.0 Algorithm. The protocol is designed to preserve the statistical information derived from the vertically partitioned data available with the parties involved and provide for privacy preservation without key distribution techniques that are generally adopted. The proposed protocol KDLPDM utilizes the benefits of the commutative RSA algorithm for over head reduction yet providing privacy. The protocol is compared with the ACP group communication protocol and its advantages are discussed in results section of this paper.

Rating: 
5
Average: 5 (1 vote)
Your rating: None