<![CDATA[The Free Nutritious Meal Program (MBG) is a public policy that requires evaluation based on public opinion. This study developed a Long Short-Term Memory (LSTM) model to classify public sentiment from 13,923 X reviews, collected using the tweet-harvest library. The data was processed with Word2Vec weighting and Lexicon-Based labeling, resulting in 73.4% positive sentiment and 26.6% negative sentiment. The model was tested with train-test split ratios of 60:40, 70:30, 80:20, and 90:10, with the best performance at a ratio of 80:20 (91.71% accuracy, 89% precision, 90% recall, 89% F1-score). The model architecture includes Embedding, LSTM (128 units), Dropout (70%), and Dense layers, optimized with categorical_crossentropy and Adam. The confusion matrix evaluation shows the effectiveness of the model, despite weak negative classes due to data imbalance. The results provide insights for improving MBG implementation, with LSTM excelling at capturing text patterns compared to SVM and BERT.]]>
Copyrights © 2025
