An efficient approach for maintaining association rules based on adjusting FP-tree structures

Abstract

In this study, a general incremental updating technique is proposed for maintaining the frequent itemsets discovered in a database in the cases including insertion, deletion, and modification of transactions in the database. An efficient algorithm, called AFPIM (Adjusting FP-tree for Incremental Mining), is designed based on adjusting FP-tree structures. Our approach uses a FP-tree structure to store the compact information of transactions involving frequent and pre-frequent items in the original database. In most cases, without needing to rescan the original database, the new FP-tree structure of the updated database can be obtained by adjusting FP-tree of the original database according to the changed transactions. Experimental results show that AFPIM outperforms the existing algorithms in terms of the execution time. © Springer-Verlag 2004.