Associative Classification to Categorize Arabic Data Sets

Associative classification (AC) is a promising data mining approach which builds more accurateclassifiers than traditional classification technique such as decision trees and rule induction. Byintegrating association rules mining with classification, AC has two main phases which are rulegeneration and classifier building.In this paper, we investigate one of the well known AC algorithm i.e. CBA, Na ve Bayesian method(NB) and Support Vector Machine algorithm (SVM) on different Arabic data sets. The bases of ourcomparison are the most popular text evaluation measures. The Experimental results against differentArabic text categorization data sets reveal that CBA algorithm outperforms the NB and SVMalgorithms with regards to all measures.