Arabic Text Classification on Full Word

Ahmed, Mahmoud and Elhassan, Rasha (2015) Arabic Text Classification on Full Word. International Journal of Computer Science and Software Engineering, 4 (5). pp. 114-120. ISSN 2409-4285

[img]
Preview
PDF
Arabic Text Classification on Full Word.pdf

Download (488kB) | Preview
Official URL | رابط موقع المجلة: http://www.IJCSSE.org

Abstract|ملخص البحث

Text classification aims to extract the useful information from a large data. The documents may contain unnecessary data which may affect the accuracy of the classifier. Data preprocessing phase aims to clean the texts by removing unnecessary information. The main object of this paper is to explain and determine the effectiveness of the data preprocessing on full word in the accuracy of both training model and classifier. This will be done by two approaches: first the observation of data set contain and second the stop word estimation technique. In the experiment, the Sequential Minimal Optimization (SMO), Naïve Bayesian (NB) J48 and K-nearest neighbors (KNN) were used to build the training models. By implement the two approaches and measured the accuracy by precision, recall and f- measure, the results showed that the SMO classifier outperforms the three other classifiers as a training model and a classifier.

Item Type|تصنيف النتاج العلمي: Article| منشور علمي
Subjects | مجال موضوع النشر: Computer
Computer Sciences
Divisions | الكلية: College of Community - Khami Mushyait > Information Systems
Depositing User: Dr. Rasha Elhassan
Date Deposited: 11 Feb 2019 09:15
Last Modified: 11 Feb 2019 09:15
URI: http://eprints.kku.edu.sa/id/eprint/3167

Actions (login required)

View Item | استعراض View Item | استعراض