Статья

Arabic Sentiment Analysis for ChatGPT Using Machine Learning Classification Algorithms: A Hyperparameter Optimization Technique

Ahmad NasayrehDepartment of Information Technology and Computer Sciences, Yarmouk University, Irbid, JordanRabia Emhamed Al MamlookDepartment of Business Administration, Trine University, Angola, IN, USA and Department of Industrial Engineering, University Zawia, Tripoli, LibyaGhassan SamaraDepartment of Computer Science, Faculty of Information Technology, Zarqa University, Zarqa, JordanHasan GharaibehDepartment of Information Technology and Computer Sciences, Yarmouk University, Irbid, JordanMohammad AljaidiDepartment of Computer Science, Faculty of Information Technology, Zarqa University, Zarqa, JordanDalia Alzu’biDepartment of Computer Science and Engineering, Concordia University, Montreal, CanadaEssam Al-DaoudDepartment of Computer Science, Faculty of Information Technology, Zarqa University, Zarqa, JordanLaith AbualigahComputer Science Department, Al al-Bayt University, Mafraq, Jordan; Hourani Center for Applied Scientific Research, Al-Ahliyya Amman University, Amman, Jordan; MEU Research Unit, Middle East University, Amman, Jordan; Department of Electrical and Computer Engineering, Lebanese American University, Byblos, Lebanon; School of Computer Sciences, Universiti Sains Malaysia, Pulau Pinang, Malaysia; School of Engineering and Technology, Sunway University Malaysia, Petaling Jaya, Malaysia; Applied Science

2024en

ABI

Аннотация

In the realm of ChatGPT's language capabilities, exploring Arabic Sentiment Analysis emerges as a crucial research focus. This study centers on ChatGPT, a popular machine learning model engaging in dialogues with users, garnering attention for its exceptional performance and widespread impact, particularly in the Arab world. The objective is to assess people's opinions about ChatGPT, categorizing them as positive or negative. Despite abundant research in English, there is a notable gap in Arabic studies. We assembled a dataset from X (formerly known as Twitter), comprising 2,247 tweets, classified by Arabic language specialists. Employing various machine learning algorithms, including Support Vector Machine (SVM), Logistic Regression (LR), Random Forest (RF), and Naïve Bayes (NB), we implemented hyperparameter optimization techniques such as Bayesian optimization, Grid Search, and random search to select the best hyperparameters that contribute to achieving the best performance. Through training and testing, performance enhancements were observed with optimization algorithms. SVM exhibited superior performance, achieving 90% accuracy, 88% precision, 95% recall, and 91% F1 score with Grid Search. These findings contribute valuable insights into ChatGPT's impact in the Arab world, offering a comprehensive understanding of sentiment analysis through machine learning methodologies.

Перевод пока недоступен

Идентификаторы

DOI: 10.1145/3638285

Цитирования и источники

Цитирований: 3Использованных источников: 0

Показатели — AkademScholar