SECURE MINING OF ASSOCIATION RULES IN DISTRIBUTED DATABASE USING SEMI HONEST THIRD PARTY

Data mining is used to discovering useful patterns hidden in a database from large datasets, but sometimes these datasets are split among various sites and none of the sites is allowed to expose its database to another site. Association Rule mining in distributed database is one of the important and well researched techniques of data mining. This technique discloses some interesting relationship between loc al as well as global item sets. Mining of association rules from distributed databasesare essential in different area such as market basket analysis . But s ometimes there are problem to determine a useful p attern in distributed databas es . Also t he protection of information from illegal access has been a long term goal for businesses and government organizations. So that i t requires enhanced privacy. In this paper, we have shown the Association rule mining algorithm over horizontal di strib uted databases. Using our approach is to generate strong association rules from different data sets spread over various geographical sites and also preserving privacy of data.