<![CDATA[The pros and cons of using ChatGPT in education have caused academic debate as it has influenced current educational praxis. Discussions about the possibility of ChatGPT for writing manuscripts or doing assignments are rife on social media, one of which is Twitter. The purpose of this study is to understand the public perception of the use of ChatGPT in education. The proposed method is sentiment analysis with SVM and Majority Voting algorithms. SVM is one of the superior algorithms in pattern recognition and is suitable for use in classification. The Majority Voting ensemble algorithm combines independent algorithms' prediction results. In this research, majority voting uses three base classifiers, namely Naïve Bayes, Random Forest, and KNN. The results of the study showed that the accuracy of SVM is 83.6% and Majority Voting is 85.4%, with the accuracy of the NB, RF, and KNN base classifiers of 76.82%, 80.91%, and 74.5%, respectively. This proved that the Majority Voting Ensemble is superior to individual algorithms with higher accuracy values. This follows the results of previous research, where the ensemble performs better than the individual algorithm. The accuracy values of SVM and the Ensemble Majority Voting models showed that both models could successfully classify sentiment on tweet data for using ChatGPT in education.]]>
Copyrights © 2025
