Analisis Sentimen Opini Publik Terhadap Cyberbullying Pada Komentar Instagram Menggunakan Multinomial Naive Bayes

This study aims to evaluate the effectiveness of a sentiment classification model using the Naïve Bayes algorithm on Instagram comment data. The main focus of this research is to measure the performance of the model in terms of accuracy, precision, recall, and F1-score. The data used in this study includes 400 Instagram comments that have been labeled with negative and positive sentiments. Data pre-processing involved case folding, tokenization, stopword removal, and stemming, followed by TF-IDF weighting to measure the importance of each word. The data was divided into 80% for training and 20% for testing. The Naïve Bayes model was then applied to the test data to predict sentiment. The evaluation results show that the model achieved an accuracy of 86.25%, with a precision of 85.56%, recall of 86.46%, and F1-score of 85.88%. For the negative class, the precision reached 91%, recall 85%, and F1-score 88%, while for the positive class, the precision was 80%, recall 88%, and F1-score 84%. These findings show that the Naïve Bayes model is effective in classifying the sentiment of Instagram comments and provides useful insights in understanding public sentiment towards the issue of cyberbullying