Saudi Arabic Multi-dialects Identification in Social Media Texts

ChatGPT is a state-of-the-art, robust artificial intelligence language model that can be used in a wide range of Natural Language Processing (NLP) applications. These applications include, but are not restricted to, text classification, text generation, sentiment analysis, and question answering. ChatGPT is primarily aimed at generating English text, but it has also been found to process other languages, such as Arabic language. This paper shows the usage of the ChatGPT model in the task of dialect identification, specifically for Saudi Arabic dialects. Five different Saudi Arabic dialects, namely Hijazi (spoken by people in western regions), Najdi (spoken by people in central regions), Eastern (spoken by people in eastern regions), Southern (spoken by people in southern regions), and Northern (spoken by people in northern regions) were selected in this study. The experimental results demonstrate that ChatGPT achieved an overall accuracy of 0.42, which is higher than identification with a Support Vector Machine (SVM) that gave 0.33 in our sample data-set.