Survey on Secure Mining of Association Rules in Vertically Distributed Databases

A distributed database system is a collection of sites connected on a common high bandwidth network. Logically, data belongs to the same system but physically it is spread over the sites of the network, making the distribution invisible to the user. The advantage of this distribution resides in achieving availability, performance, modularity and reliability. In this paper, I have done a survey of papers related to Mining of Association Rules over distributed databases. From this survey, we have come up with a proposed solution to address the problem of secure mining of association rules where transactions are distributed in vertically distributed databases. Each site holds some attributes of each transaction and the sites wish to participate in the identification of globally valid association rules However, the sites should not reveal individual transaction data. The Protocol is based on Apriori Algorithm [2] and MultiParty Algorithm [3] for efficiently discovering frequent item sets with minimum support levels, without either site communicating individual transaction values. DOI: 10.17762/ijritcc2321-8169.15035