Frequent Patterns Mining using only one-time Database Scan


The KIPS Transactions:PartD, Vol. 15, No. 1, pp. 15-30, Feb. 2008
10.3745/KIPSTD.2008.15.1.15,   PDF Download:

Abstract

In this paper, we propose an efficient algorithm using only one-time database scan. The proposed algorithm creates the bipartite graph which indicates relationship of large items and transactions including the large items. And then we can find large itemsets using the bipartite graph. The bipartite graph is generated when database is scanned to find large items. We can't easily find transactions which include large items in the large database. In the bipartite graph, large items and transactions are linked each other. So, we can trace the transactions which include large items through the link information. Therefore the bipartite graph is a indexed database which indicates inclusion relationship of large items and transactions. We can fast find large itemsets because proposed method conducts only one-time database scan and scans indexed the bipartite graph. Also, it don't generate candidate itemsets.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
D. J. Chai, L. Jin, Y. M. Lee, B. H. Hwang, K. H. Ryu, "Frequent Patterns Mining using only one-time Database Scan," The KIPS Transactions:PartD, vol. 15, no. 1, pp. 15-30, 2008. DOI: 10.3745/KIPSTD.2008.15.1.15.

[ACM Style]
Duck Jin Chai, Long Jin, Yong Mi Lee, Bu Hyun Hwang, and Keun Ho Ryu. 2008. Frequent Patterns Mining using only one-time Database Scan. The KIPS Transactions:PartD, 15, 1, (2008), 15-30. DOI: 10.3745/KIPSTD.2008.15.1.15.