Preservation of the Privacy for Multiple Custodian Systems with Rule Sharing

The aim of the research work is to mine the data set available with each custodian in a semi honest model, securely without disclosure of any data amongst various custodians involved. No custodian discloses any information. In the proposed scheme in order to reduce the computational complexity, the data partitioning has been done in the horizontal way. The proposed research work consists of a well skilled and planned architecture implementation for achieving the proposed privacy preservation in the data mining filed and used a new hybrid data mining model which is developed for combining commutative RSA and a C5.0 algorithm to generate classification rules. This study utilized real world data collected from an UCI repository and experiments are conducted based on the parameters like time complexity, accuracy and error rate. The proposed model preserve expected level of privacy without any information loss, take less time for computation, lower error rate and improves accuracy.