Mining multi-level rules with recurrent items using FP'-Tree

• Main Authors: ONG, Kok-Leong, NG, Wee-Keong, LIM, Ee Peng
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2001
• Subjects: Databases and Information Systems; Numerical Analysis and Scientific Computing
• Online Access: https://ink.library.smu.edu.sg/sis_research/904; https://ink.library.smu.edu.sg/context/sis_research/article/1903/viewcontent/ong01mining.pdf
• Institution: Singapore Management University

Association rule mining has received broad research in the academic and wide application in the real world. As a result, many variations exist and one such variant is the mining of multi-level rules. The mining of multi-level rules has proved to be useful in discovering important knowledge that conventional algorithms such as Apriori, SETM, DIC etc., miss. However, existing techniques for mining multi-level rules have failed to take into account the recurrence relationship that can occur in a transaction during the translation of an atomic item to a higher level representation. As a result, rules containing recurrent items go unnoticed. In this paper, we consider the notion of `quantity' to an item, and present an algorithm based on an extension of the FP-Tree to find association rules with recurrent items at multiple concept levels.