Arabic Sentiment Analysis for Student Evaluation using Machine Learning and the AraBERT Transformer

Abstract

Recently, Sentiment Analysis (SA) has become a crucial area of research as it enables us to gauge people's opinions from various sources such as student evaluations, social media posts, product reviews, etc. This paper aims to create an Arabic dataset derived from student satisfaction surveys conducted at the University of Jeddah regarding their subjects and instructors. In addition, this study presents an evaluation of classical machine learning models such as Naive Bayes, Support Vector Machine, Logistic Regression, Decision Tree, and Random Forest classifier for Arabic SA, whereas the results are compared using various metrics. Furthermore, AraBERT was used for the pre-trained transformer to improve the performance, achieving an accuracy of 78%. The paper fills the lack of SA research in the education domain in the Arabic language.