Securely mining transactional databases for association rules using FDM

Data are very significant in various ways. Data are facts or statistics collected together for reference or analysis. As the data are growing continuously, there should be means to find the correlations among them and use those relationships for analysis and prediction. This paper presents how horizontally distributed databases are mined for association rules using Fast Distributed Mining. These association rules are securely mined to get the anonymized view of the data that can be used for prediction. The databases have common attributes and databases are partitioned on those common attributes. Partitioning of large databases horizontally will result in data sets that are smaller and manageable. The efficiency will increase in terms of query execution. With horizontal partitioning, the columns will not be partitioned but rows will be split based on certain criteria so as to minimize querying across multiple partitions.