A Rough-Apriori Technique in Mining Linguistic Association Rules

This paper has proposed a rough-Apriori based mining technique in mining linguistic association rules focusing on the problem of capturing the numerical interval with linguistic terms in quantitative association rules mining. It uses the rough membership function to capture the linguistic interval b...

Full description

Bibliographic Details
Main Authors: Choo, Yun Huoy, Abu Bakar, Azuraliza, Hamdan, Abdul Razak
Format: Book Chapter
Language:English
Published: Springer Berlin Heidelberg 2008
Subjects:
Online Access:http://eprints.utem.edu.my/id/eprint/151/
http://eprints.utem.edu.my/id/eprint/151/1/ARoughAprioriTechniqueInMiningLinguisticAR.pdf
Description
Summary:This paper has proposed a rough-Apriori based mining technique in mining linguistic association rules focusing on the problem of capturing the numerical interval with linguistic terms in quantitative association rules mining. It uses the rough membership function to capture the linguistic interval before implementing the Apriori algorithm to mine interesting association rules. The performance of conventional quantitative association rules mining algorithm with Boolean reasoning as the discretization method was compared to the proposed technique and the fuzzy-based technique. Five UCI datasets were tested in the 10-fold cross validation experiment settings. The frequent itemsets discovery in the Apriori algorithm was constrained to five iterations comparing to maximum iterations. Results show that the proposed technique has performed comparatively well by generating more specific rules as compared to the other techniques.