Arabic Text Classification Using Maximum Weight Algorithm

Abstract

Many algorithms of classification implemented to the issue of text categorization.A large portion of the work implemented in the English text. On the otherhand, very few researchers implemented in the Arabic text. The nature of Arabictext is very different than English text, and the preprocessing of the Arabictext is extremely difficult and more challenging. In this paper, Maximum Weight(MW) algorithm applied after preprocessing on the Arabic dataset that consists of(16757) text files used for the first time. The results showed that MW is applicableto Arabic text, it reached about (0.83) on average. 10-fold cross-validation used tothe reliability of the result.