Leveraging Machine Learning Techniques to Analyze Consumer Mindset Metrics Embedded in Arabic Dialect Texts Across Social Media Platforms

Abstract

As social media grows in popularity around the world, analyzing Arabic texts on these platforms can provide important insights into consumer attitudes and behavior. The complexity and diversity of Arabic and its dialects, however, make it a challenging task. This research raises these challenges by using and comparing the performance of Machine Learning (ML) models for classifying social media comments in Arabic into service quality, loyalty, purchase intention, and satisfaction types. This research employed several machine learning models, including Support Vector Machines (SVM), Multinomial Naïve Bayes, Linear Support Vector Classifier (SVC), and K-Nearest Neighbors (KNN). The results indicate that the Linear SVC outperforms the other models and represents the most effective approach. Furthermore, the classifiers demonstrate strong performance in Arabic short text classification, confirming the effectiveness of machine learning techniques in extracting meaningful insights from Jordanian dialect social media comments.