<![CDATA[Cyberbullying is aggressive behavior through electronic media that has a serious psychological impact on victims, especially artists who are often the target of negative comments that can influence the behavior of other users on social media. This action is not only psychologically and mentally damaging, but is also a cyber crime that needs to be followed up. This research aims to identify and predict negative comments that lead to cyberbullying through text mining with classification and sentiment analysis methods. This study compares two classification methods: Naive Bayes and Decision Tree, to determine which method is more accurate. Data was taken from 1680 comments on Indonesian artists' Instagram accounts, from September 2023 to May 2024, and divided into 80% for training and 20% for testing, so that the results obtained were Naive Bayes showing 80.95% accuracy, 80.95% precision, recall 87.46%, and F1 score 84.02%, while Decision Tree shows accuracy 82.44%, precision 81.68%, recall 82.44%, and F1 score 81.86%. The findings show that Decision Tree has higher accuracy in classifying cyberbullying comments than Naive Bayes.]]>
Copyrights © 2024
