Alternative Method for Increnentally Constructing the FP-Tree

The FP-tree is an effective data structure that facilitates the mining of frequent patterns from transactional databases. But, transactional databases are dynamic in general, and hence modifications on the database must be reflecting onto the FP-tree. Constructing the FP-tree from scratch and incrementally updating the FP-tree are two possible choices. However, from scratch construction turns unfeasible as the database size increases. So, this paper addresses incremental update by extending the FP-tree concepts and manipulation process. Our new approach is capable of handling all kinds of changes; include additions, deletions and modifications. The target is achieved by constructing and incrementally dealing with the complete FP-tree, i.e., with one minimum support. Constructing the complete FP-tree has the other advantage that it provides the freedom of mining for lower minimum support values without the need to reconstruct the tree. However, directly reflecting the changes onto the FP-tree may invalidate the basic FP-tree structure. Thus, we apply a sequence of shuffling and merging operations to validate and maintain the modified tree. The experiments conducted on synthetic and real datasets clearly highlight advantages of the proposed incremental approach over constructing the FP-tree from scratch

[1]  Rakesh Agarwal,et al.  Fast Algorithms for Mining Association Rules , 1994, VLDB 1994.

[2]  Jia-Ling Koh,et al.  An Efficient Approach for Maintaining Association Rules Based on Adjusting FP-Tree Structures1 , 2004, DASFAA.

[3]  David Wai-Lok Cheung,et al.  A General Incremental Technique for Maintaining Discovered Association Rules , 1997, DASFAA.

[4]  Reda Alhajj,et al.  Constructing Complete FP-Tree for Incremental Mining of Frequent Patterns in Dynamic Databases , 2006, IEA/AIE.

[5]  Wee Keong Ng,et al.  Rapid association rule mining , 2001, CIKM '01.

[6]  Jian Pei,et al.  Mining frequent patterns without candidate generation , 2000, SIGMOD '00.

[7]  Sanjay Ranka,et al.  An Efficient Algorithm for the Incremental Updation of Association Rules in Large Databases , 1997, KDD.

[8]  Amihood Amir,et al.  A New and Versatile Method for Association Generation , 1997, Inf. Syst..

[9]  David Wai-Lok Cheung,et al.  Maintenance of Discovered Knowledge: A Case in Multi-Level Association Rules , 1996, KDD.

[10]  Johannes Gehrke,et al.  DEMON: Mining and Monitoring Evolving Data , 2001, IEEE Trans. Knowl. Data Eng..

[11]  Yonatan Aumann,et al.  Efficient Algorithms for Discovering Frequent Sets in Incremental Databases , 1997, DMKD.

[12]  R. Alhajj,et al.  Performance analysis of incremental update of association rules mining approaches , 2005, 2005 IEEE International Conference on Intelligent Engineering Systems, 2005. INES '05..

[13]  Jiawei Han,et al.  Maintenance of discovered association rules in large databases: an incremental updating technique , 1996, Proceedings of the Twelfth International Conference on Data Engineering.

[14]  Tomasz Imielinski,et al.  Mining association rules between sets of items in large databases , 1993, SIGMOD Conference.

[15]  Necip Fazil Ayan,et al.  An efficient algorithm to update large itemsets with early pruning , 1999, KDD '99.