Comparative evaluation of text classification techniques using a large diverse Arabic dataset

• Published in: Language Resources and Evaluation Vol. 47; no. 2; pp. 513 - 538
• Main Authors: Khorsheed, Mohammad S., Al-Thubaity, Abdulmohsen O.
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer 01.06.2013; Springer Netherlands; Springer Nature B.V
• Subjects: Algorithms; Applied linguistics; Arabic language; Benchmarking; Boolean data; Classification; Computational Linguistics; Computer Science; Data visualization; Datasets; Decision trees; English language; Experimentation; Human; Information classification; Language and Literature; Linguistics; Literature reviews; News content; Original Paper; Poetry; Religious poetry; Social Sciences; Support vector machines; Texts; Training; Weighting methods; Words; Writers; Arabic text categorization; Machine learning; Arabic text classification; Automatic classification; Arabic; Text; Natural language processing; Algorithm; Categorization
• ISSN: 1574-020X; 1572-8412; 1574-0218
• DOI: 10.1007/s10579-013-9221-8

A vast amount of valuable human knowledge is recorded in documents. The rapid growth in the number of machine-readable documents for public or private access necessitates the use of automatic text classification. While a lot of effort has been put into Western languages—mostly English—minimal experimentation has been done with Arabic. This paper presents, first, an up-to-date review of the work done in the field of Arabic text classification and, second, a large and diverse dataset that can be used for benchmarking Arabic text classification algorithms. The different techniques derived from the literature review are illustrated by their application to the proposed dataset. The results of various feature selections, weighting methods, and classification algorithms show, on average, the superiority of support vector machine, followed by the decision tree algorithm (C4.5) and Naïve Bayes. The best classification accuracy was 97 % for the Islamic Topics dataset, and the least accurate was 61 % for the Arabic Poems dataset.