Mining Regular Patterns in Transactional Databases

Syed Khairuzzaman TANBEER  Chowdhury Farhan AHMED  Byeong-Soo JEONG  Young-Koo LEE  

IEICE TRANSACTIONS on Information and Systems   Vol.E91-D   No.11   pp.2568-2577
Publication Date: 2008/11/01
Online ISSN: 1745-1361
DOI: 10.1093/ietisy/e91-d.11.2568
Print ISSN: 0916-8532
Type of Manuscript: Special Section PAPER (Special Section on Knowledge, Information and Creativity Support System)
Category: Knowledge Discovery and Data Mining
data mining,  pattern mining,  regular pattern,  periodic pattern,  

Full Text: PDF(585.4KB)>>
Buy this Article

The frequency of a pattern may not be a sufficient criterion for identifying meaningful patterns in a database. The temporal regularity of a pattern can be another key criterion for assessing the importance of a pattern in several applications. A pattern can be said regular if it appears at a regular user-defined interval in the database. Even though there have been some efforts to discover periodic patterns in time-series and sequential data, none of the existing studies have provided an appropriate method for discovering the patterns that occur regularly in a transactional database. Therefore, in this paper, we introduce a novel concept of mining regular patterns from transactional databases. We also devise an efficient tree-based data structure, called a Regular Pattern tree (RP-tree in short), that captures the database contents in a highly compact manner and enables a pattern growth-based mining technique to generate the complete set of regular patterns in a database for a user-defined regularity threshold. Our performance study shows that mining regular patterns with an RP-tree is time and memory efficient, as well as highly scalable.