An Experimental Study for the Effect of Stop Words Elimination for Arabic Text Classification Algorithms

An Experimental Study for the Effect of Stop Words Elimination for Arabic Text Classification Algorithms

Bassam Al-Shargabi (Isra University, Jordan), Fekry Olayah (Isra University, Jordan) and Waseem AL Romimah (University of Science and Technology, Yemen)
DOI: 10.4018/978-1-4666-2157-2.ch012
OnDemand PDF Download:
List Price: $37.50


In this paper, an experimental study was conducted on three techniques for Arabic text classification. These techniques are Support Vector Machine (SVM) with Sequential Minimal Optimization (SMO), Naïve Bayesian (NB), and J48. The paper assesses the accuracy for each classifier and determines which classifier is more accurate for Arabic text classification based on stop words elimination. The accuracy for each classifier is measured by Percentage split method (holdout), and K-fold cross validation methods, along with the time needed to classify Arabic text. The results show that the SMO classifier achieves the highest accuracy and the lowest error rate, and shows that the time needed to build the SMO model is much lower compared to other classification techniques.
Chapter Preview

Preprocessing And Classification Process

The preprocessing of data set deals with the elimination of non-meaningful words which do not indicate semantic content of the document. Some words appear in the sentences and don’t have any meaning or indications about the content such as (so لذلك, with بالإشارة, confirmation تأكيدا, for بالنسبة) or appearing frequently in the document like pronouns such as (heهو, she هي, they هم). Although the prepositions like (from من, to إلى, in في, about عن) or demonstratives like (this هذا, these هؤلاء, there أولئك) or interrogatives like (where أين, which أي, who من). These words may have a bad effect on statistical information and co-occurrence of the words as stated in Abo Alkhair (2006) and Said, Wanas, Darwish, and Hegazy (2009).

Also the numbers and symbols like (@, #, &, %, *) and some words that indicates a sequence of the sentences like (firstly أولا, secondly ثانيا, thirdly ثالثا), these words will be considered as an Arabic stop words. Some Arabic documents may contain foreign words, especially science documents, these words are alo considered as stop words as in Al-Shalabi, Kanaan, Jaam, Hasnah, and Hilat (2004) and El-Kourdi, Bensaid and Rachidi (2004).

Complete Chapter List

Search this Book: