Analisis Sentimen Program Makan Siang Gratis Menggunakan Multinomial Naive Bayes

Twitter is one of the social media networks used by the public to express opinions, criticisms and points of view. A topic that has been widely discussed is the free lunch work program promoted by the presidential candidate pair Prabowo and Gibran. The existence of pros and cons in assessing a policy or work program is very high, an approach is needed to analyze public sentiment towards the program. Analysis of public sentiment towards this program is important to provide an overview of how well the program is received by the public and how public opinion affects the program. This research aims to determine the sentiment category towards the free lunch program using the Naive Bayes method. This research involves collecting and analyzing tweets related to “Free Lunch Program” from Twitter(X), using authentication tokens. The data was processed through pre-processing, then classified with Multinomial Naive Bayes. A total of 1902 data were obtained from labeling with the Lexicon Based method. The results obtained 84.06% accuracy, 83.9% precision, 98.9% recall, and 90.70% F1-score calculated using confusion matrix. The sentiment analysis results show that the majority of community responses tend to be positive, in other words, the community supports and is optimistic about the free lunch work program which can directly provide social benefits to the community in fulfilling basic needs