Classification Arabic language (Classical Arabic Poetry, Al-Hur Arabic Poetry and Prose) Using Machine Learning

Abstract

Many languages globally have made significant advances in electronically studying and classifying texts. Making electronic text a great alternative to manual classification by saving time, cost, and effort. However, Arabic has not seen similar progress due to several limitations faced by researchers, such as the complexity of the language, the scarcity of related research, and the use of classical Arabic. Additionally, the poetry presents further challenges, like the reliance on a single activation function. This paper introduces a new method for classifying the Arabic text (Prose, classical Arabic poetry and Al-hur Arabic poetry) based on distinctive features that identify the type of Arabic text. Prepressing data is crucial in this approach as it enhances classification accuracy.