<![CDATA[The viral spread of a video in April 2026 featuring electric motorcycles branded with the National Nutrition Agency (Badan Gizi Nasional) logo sparked public debate on X regarding the budget transparency of the Free Nutritious Meal (MBG) Program, specifically concerning the procurement of electric motorcycles for the Nutrition Fulfillment Service Units (SPPG). The sheer volume of data and the rapid dissemination of information on social media render manual analysis inadequate. This study aims to map the distribution of public sentiment (positive, neutral, negative), measure the performance of the classification model, and identify dominant keywords within each sentiment category. Data were collected via web scraping from X using keywords related to the SPPG electric motorcycle procurement issue during the April–May 2026 period. Following preprocessing, automated labeling was performed using the IndoBERT model with a confidence threshold of 0.60, and the results were manually validated against a sample of 500 data points. Feature extraction utilized TF-IDF, while sentiment classification employed a Support Vector Machine (SVM) algorithm based on LinearSVC, evaluated using Stratified 5-Fold Cross-Validation. The results reveal a predominance of neutral sentiment (43.25%), followed by negative sentiment (38.73%) and positive sentiment (18.01%). The SVM model achieved an accuracy of 78.18% with a macro F1-score of 0.77. Data from May 2026 indicate a rise in negative sentiment to 66.01%, signaling widespread public concern regarding the transparency and budget governance of the MBG program.]]>
Copyrights © 2026
