Abstract:

Data mining refers to extracting or mining knowledge from large amount of data. It helps in discovering the frequent patterns. This knowledge can be help in making business decisions conventional data mining techniques were proposed to be in efficient, the source of data is incremental in nature and also they need to be re executed every time which in turn multiple database scans.

In this paper, we propose a novel data structure called CATS Tree to suit for continuously growing data sets, based on association rule mining with incremental techniques. CATS Tree extends the idea of FP-Tree to improve storage compression and allow frequent pattern mining without generation of candidate itemsets. The proposed algorithms enable frequent pattern mining with different supports without rebuilding the tree structure. Furthermore, the algorithms allow mining with a single pass over the database as well as efficient insertion or deletion of transactions at any time.

This paper introduces Compressed and Arranged Transaction Sequences Tree or CATS Tree and CATS Tree algorithms. Once CATS Tree is built, it can be used for multiple frequent pattern mining with different supports. Furthermore, CATS Tree and CATS Tree algorithms allow single pass frequent pattern mining and transaction stream mining. In addition, transactions can be added to or removed from the tree at any time.It is assumed that there is no limitation on the main memory. The assumption is realistic for a reasonably large database due to the following reasons: 1) the current trend of modern computing moves towards computers with large amounts of main memory (gigabytes sized); 2) memory management techniques and data compression technique in the CATS Tree reduce memory footprint; 3) the same assumption has been used in other publications . In addition, CATS Tree allows removal of transactions concurrently. Even a huge database can be processed by CATS Tree if out-of-date transactions are removed concurrently.