Menjelang Pemilihan Presiden 2024, pasangan calon presiden dan wakil presiden Prabowo Subianto – Gibran Rakabuming Raka menarik perhatian masyarakat dengan elektabilitas tinggi. Penelitian ini menganalisis sentimen masyarakat terhadap pasangan Prabowo – Gibran di Twitter (X) menggunakan metode Naive Bayes. Tahap pre processing meliputi cleansing, case folding, normalisasi, tokenizing, filtering, stemming. Hasil pelabelan menunjukkan bahwa dari 1063 data, terdapat 443 data sentimen positif, 343 data sentimen netral, dan 277 sentimen negatif. Data dibagi menjadi data asli dan data yang telah di-SMOTE, dengan 90% sebagai data latih dan 10% sebagai data uji. Berdasarkan hasil penelitian, evaluasi menunjukkan akurasi model Naive Bayes mencapai 73% untuk data asli dan 75% untuk data yang telah di-SMOTE yang menunjukan adanya peningkatan akurasi dengan penerapan teknik SMOTE menggunakan rasio 90:10.

Ahead of the 2024 Presidential Election, the presidential and vice-presidential candidates Prabowo Subianto – Gibran Rakabuming Raka are drawing significant public attention due to their high electability. This study analyzes public sentiment towards the Prabowo – Gibran candidates on Twitter (X) using the Naive Bayes method. The pre-processing stage includes cleansing, case folding, normalization, tokenizing, filtering, and stemming. The labeling results indicate that out of 1063 data points, there are 443 positive sentiment data points, 343 neutral sentiment data points, and 277 negative sentiment data points. The data is divided into original data and data that has been SMOTE-applied, with 90% used for training and 10% for testing. The evaluation results show that the Naive Bayes model achieves 73% accuracy for the original data and 75% accuracy for the SMOTE applied data, demonstrating an improvement in accuracy with the application of the SMOTE technique using a 90:10 ratio.