Sentiment Analysis of the Free Nutritious Meal (MBG) Program on TikTok Using Naïve Bayes Classifier

Abstract

The Free Nutritious Meal (MBG) Program is an Indonesian government initiative to improve nutritional quality and support human resource development. As a public policy, it has generated diverse public responses on social media, particularly TikTok. The novelty of this study lies in utilizing TikTok comments, which remain underexplored in previous MBG sentiment studies, and employing expert-validated manual sentiment labeling to improve annotation reliability. This study applies the Naïve Bayes Classifier to analyze public sentiment toward the MBG Program using 3,190 TikTok comments, of which 1,294 were retained after relevance filtering, data cleaning, and manual labeling validated by two Indonesian language experts. Text preprocessing included cleaning, case folding, tokenization, normalization, stopword removal, and stemming, followed by TF–IDF feature weighting. The model was evaluated using stratified 10-fold cross-validation. It achieved 72.80% accuracy, 69.90% weighted precision, 73.20% weighted recall, and 71.40% F1-score. Negative sentiment dominated public discussion of the MBG Program. These findings demonstrate that combining expert-validated manual labeling, TF–IDF, and the Naïve Bayes Classifier provides a reliable approach for sentiment analysis of Indonesian social media text and offers useful insights for evaluating the implementation of the MBG Program