A EFFECTIVE METHOD FOR FREQUENT ITEMSETS MINING ON VERY LARGE TRANSACTION DATABASE ENVIRONMENT

摘要

All of this invention is a set of dense and sparse frequent topic frequent item sets in large transactional database environment relates to a method of. The step of configuring the header table this invention is to count all the items in the transaction in the source database transaction; The Patricia Frequent Pattern tree is compressed using a configured header table (Compressed Patricia Frequent Patterns Tree: CPFP-Tree) steps to configure; The step of configuring the individual CPFP-Tree for each item using the configured CPFP-Tree and the header table; And the constructed using individual CPFP-Tree by yirueojim including the step of searching the frequent item sets, and propose a new data structure CPFP-Tree at the same time, both dense and sparse frequent item set of frequent item sets quickly and efficiently allowing you to navigate.