Basic Framework of CATSIM Tree, for Efficient Frequent Pattern Mining

Finding frequent patterns from databases have been the most time consuming process inassociation rule mining. Several effective data structures, such as two-dimensional arrays, graphs, treesand tries have been proposed to collect candidate itemsets and frequent itemsets. It seems that thetree structure is most extractive to storing itemsets. The outstanding tree has been proposed so far iscalled FP-tree which is a prefix tree structure. Some advancement with this tree structure is calledCATS tree. CATS Tree extends an idea of FP-Tree to improve storage compression and allow frequentpattern mining without generation of candidate itemsets. It allows the mining with a single pass overthe database. In this work, CATSIM Tree is presented for which an attempt has been made to modifypresent CATS Tree in order to make it efficient for incremental mining