<![CDATA[The Free Lunch Program is a government policy designed to improve the nutritional status and health of students at the primary and secondary education levels. The initiative has generated diverse public responses, particularly on social media, where Twitter functions as a major platform for rapid and open opinion exchange. Therefore, this study aims to analyze public sentiment toward the Free Lunch Program using a machine-learning approach to obtain a quantitative overview of public perceptions. The research process began with data collection using Tweet-Harvest, followed by a preprocessing stage that included text cleaning, normalization, tokenization, and stop-word removal. Afterward, sentiment labeling and class balancing were performed to reduce bias, and feature extraction using Term Frequency–Inverse Document Frequency (TF-IDF) was applied to generate informative text representations. A classification model was then constructed using the Multinomial Naïve Bayes algorithm because of its suitability for high-dimensional textual data. The evaluation results showed that the model achieved an accuracy of 61%, with the highest performance recorded in the positive sentiment class compared with neutral and negative classes. In addition, this study produced a Streamlit-based web application capable of performing automatic and interactive sentiment analysis. These findings indicate that the public tends to view the Free Lunch Program positively, while also demonstrating the effectiveness of machine-learning methods for mapping opinion through social-media text data.]]>
Copyrights © 2025
