Program Makan Bergizi Gratis merupakan salah satu program pemerintah yang
bertujuan meningkatkan kualitas gizi masyarakat, khususnya bagi anak sekolah dan
kelompok sasaran lainnya. Pelaksanaan program ini menimbulkan berbagai
tanggapan dari masyarakat yang banyak disampaikan melalui media sosial X, baik
berupa sentimen positif maupun negatif. Besarnya jumlah data opini yang tersedia
menjadikan analisis secara manual kurang efektif sehingga diperlukan metode
analisis sentimen yang mampu mengolah data teks secara otomatis. Penelitian ini
bertujuan untuk menganalisis sentimen masyarakat terhadap Program Makan
Bergizi Gratis berdasarkan data ulasan pada media sosial X menggunakan metode
Naïve Bayes Classifier serta mengevaluasi kinerja model menggunakan metrik
accuracy, precision, recall, dan F1-score.Penelitian diawali dengan pengumpulan
data menggunakan teknik web crawling melalui Tweet Harvest. Data yang
diperoleh kemudian diproses melalui tahapan preprocessing yang meliputi
cleaning, case folding, tokenizing, stemming, dan stopword removal. Selanjutnya
dilakukan pembobotan kata menggunakan metode Term Frequency–Inverse
Document Frequency (TF-IDF) sebelum dilakukan proses klasifikasi menggunakan
algoritma Naïve Bayes Classifier. Pengujian model dilakukan dengan lima skenario
pembagian data, yaitu 90:10, 80:20, 70:30, 60:40, dan 50:50. Hasil penelitian
menunjukkan bahwa metode Naïve Bayes Classifier mampu mengklasifikasikan
sentimen masyarakat dengan performa yang baik dan konsisten pada seluruh
skenario pengujian. Skenario pembagian data 90:10 menghasilkan performa terbaik
dengan nilai accuracy sebesar 87,00%, precision sebesar 86,73%, recall sebesar
100%, dan F1-score sebesar 92,89%. Hasil tersebut menunjukkan bahwa semakin
besar proporsi data latih yang digunakan, semakin baik kemampuan model dalam
mengenali pola sentimen dan menghasilkan klasifikasi yang akurat. Dengan
demikian, metode Naïve Bayes Classifier efektif diterapkan untuk analisis sentimen
masyarakat terhadap Program Makan Bergizi Gratis pada media sosial X.
Kata Kunci: Analisis Sentimen, Naïve Bayes Classifier, TF-IDF, Media Sosial X,
Program Makan Bergizi Gratis.

The Free Nutritious Meal Program is one of the Indonesian government's
initiatives aimed at improving the nutritional quality of the community, particularly
among school children and other target groups. The implementation of this program
has generated various public responses, which are widely expressed through the
social media platform X in the form of both positive and negative sentiments. The
large volume of opinion data makes manual analysis inefficient, thereby requiring
an automated sentiment analysis approach capable of processing textual data
effectively. This study aims to analyze public sentiment toward the Free Nutritious
Meal Program based on reviews collected from the social media platform X using
the Naïve Bayes Classifier method and to evaluate the model's performance using
accuracy, precision, recall, and F1-score metrics. The research began with data
collection through web crawling using Tweet Harvest. The collected data were then
processed through several preprocessing stages, including cleaning, case folding,
tokenization, stemming, and stopword removal. Furthermore, the text data were
weighted using the Term Frequency–Inverse Document Frequency (TF-IDF)
method before being classified using the Naïve Bayes Classifier algorithm. The
model was evaluated using five data-splitting scenarios: 90:10, 80:20, 70:30, 60:40,
and 50:50. The results indicate that the Naïve Bayes Classifier was able to classify
public sentiment with good and consistent performance across all testing scenarios.
The best performance was achieved using the 90:10 data-splitting scenario,
resulting in an accuracy of 87.00%, a precision of 86.73%, a recall of 100%, and an
F1-score of 92.89%. These findings indicate that a larger proportion of training data
improves the model's ability to learn sentiment patterns and produce more accurate
classifications. Therefore, the Naïve Bayes Classifier can be considered an effective
method for analyzing public sentiment toward the Free Nutritious Meal Program
on the social media platform X.
Keywords: Sentiment Analysis, Naïve Bayes Classifier, TF-IDF, Social Media X,
Free Nutritious Meal Program.