1.
5th International Conference on Natural Language and Speech Processing, ICNLSP 2022
; : 251-257, 2022.
Article
in English
| Scopus | ID: covidwho-2291096
ABSTRACT
In view of the recent interest of Saudi banks in customers' opinions through social media, our research aims to capture the sentiments of bank users on Twitter. Thus, we collected and manually annotated more than 12, 000 Saudi dialect tweets, and then we conducted experiments on machine learning models including: Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (RL) as well as state-of-the-art language models (i.e. MarBERT) to provide baselines. Results show that the accuracy in SVM, LR, RF, and MarBERT achieved 82.4%, 82%, 81%, and 82.1% respectively. Our models code and dataset will be made publicly available on GitHub. © ICNLSP 2022.All rights reserved