Privacy-Preserving Two-Party Distributed Association Rules Mining on Horizontally Partitioned Data

The need of data mining in various fields has been increased with the3 growth in storage of data. Searching the specific data is a tedious task if a wrong method is chosen. Many current applications, the data is stored in diverse locations, called as Distributed scenario. And in such scenarios, the data mining must be done for various reasons. Simultaneously, the security is very important. Opening the data for multiple users will make it more vulnerable to attacks. The misuse of the data is prime concern of the data owners within the distributed data scenario. Therefore, they wont donate their data for mining purpose and if the data is enclosed with some perceptive information, the data sharing goes low. The Privacy-Preserving data mining (PPDM) endeavors to protect the privacy of the data in the course of data mining. More effective and efficient protocols are needed because of the current generic protocols have no practical implementation in this type of case. If the participating sites are more than two, a protocol is required to be developed for the horizontally partitioned data situations occurred. Because of this reason, the Privacy Preserving distributed association rules mining protocols were developed. Yet, they tend to be dependent on a secured multi-party summary and unification computation. But these will not guarantee the security in the case of two participating parties. The implementation of the protocols for the privacy-preserving two-party distributed mining of association rule mining, a secure division computation protocols and use the commutative encryption as the foundation approach has been proposed in this paper.