KLASIFIKASI CYBERBULLYING PADA KOMENTAR VIDEO YOUTUBE MENGGUNAKAN METODE RANDOM FOREST

Abstract

Among various social media platforms, Youtube is currently one of the main destinations for young people looking for a variety of content in the form of videos. Many people are interested in content created by content creators starting from content reviewing goods and food, daily activity content, educational content, and other content, however, from the many benefits that can be obtained from the videos they make, there are still many users who do not understand ethics. in using social media which makes social media a means to intimidate other people behind the scenes which are commonly called Cyberbullying. Cyberbullying is a negative action carried out by a specific person or group by sending text messages, photos, meme images, and videos to someone's social media account to satirize, insult, harass, and discriminate against individuals. This study uses the Random Forest algorithm to predict whether a text comment on a video comment on Youtube is Bullying or not Bullying. From this study, the best value obtained from the Random forest algorithm was found in a combination of hyperparameter tuning with parameter n_estimator = 100 where the accuracy obtained was 86% and was calculated based on the evaluation results of the confusion matrix. In addition, the model built has also been able to predict new comments with fairly good predictive results, although there are still predictions that are not as they should be.