An Enhanced CART Algorithm for Preserving Privacy of Distributed Data and Provide Access Control over Tree Data

Abstract

Now in these days the utilization of distributed applications are increases rapidly because these applications are serve more than one client at a time. In the use of distributed database data distribution and management is a key area of attraction. Because of privacy of private data organizations are unwilling to participate for data mining due to the data leakage. So it is required to collect data from different parties in a secured way. This paper represents how CART algorithm can be used for multi parties in vertically partitioned environment. In order to solve the privacy and security issues the proposed model incorporates the server side random key generation and key distribution. Finally the performance of proposed classification technique is evaluated in terms of memory consumption, training time, search time, accuracy and there error rate.