Naive Bayes Classifier Based Arabic Document Categorization

Abstract

Text Categorization aims to assign an electronic document to one or more categories based on its contents. Due to the rapid growth of the number of online Arabic documents, the information libraries and Arabic document corpus, automatic Arabic document classification becomes an important task. This paper suggests the use of rooting algorithm with Nai¿ve Bayes Classifier to the problem of document categorization of Arabic language and reports the algorithm performance in terms of error rate, accuracy, and micro-average recall measures. Our experimental study shows that using rooting algorithm with Nai¿ve Bayes (NB) Classifier gives ~62.23% average accuracy and decreases the dimensionality of the training documents.