Cluster Based Partition Approach for Mining Frequent Itemsets

摘要

Data Mining is the process of extracting interesting and previously unknown patterns and correlations form huge data stored in databases. Association rule mining- a descriptive mining technique of data mining, is the process of discovering items or literals which tend to occur together in transactions. As the data to be mined is large, the time taken for accessing data is considerable. This paper describes a new approach for association mining, based on Master-Slave architecture. It uses hybrid approach - a combination of bottom up and top down approaches for searching frequent itemsets. The Apriori algorithm performs well only when the frequent itemsets are short. Algorithms with top down approach are suitable for long frequent itemsets. This new master slave architecture based algorithm combines both bottom-up and top-down approach. The Prime number based representation consumes less memory as each transaction is replaced with the product of the assigned prime numbers of their items. It reduces the time taken to determine the support count of the itemsets. The Prime number based representation offers the flexibility for testing the validity of metarules and provides reduction in the data complexity.