<![CDATA[Abstrak. BTS saat ini sedang menjalani masa hiatus karena beberapa anggota memenuhi kewajiban wajib militer di Korea Selatan. Meski tidak aktif secara grup, pencapaian individu dan kolaborasi para anggota tetap menarik perhatian. Namun, isu negatif yang beredar di media sosial berpotensi memengaruhi pandangan publik terhadap grup ini. Penelitian ini bertujuan menganalisis sentimen pengguna Twitter terhadap BTS selama masa hiatus dengan algoritma Naïve Bayes, yang efektif untuk analisis sentimen teks. Data dikumpulkan menggunakan teknik crawling pada tweet terkait BTS selama Mei–Oktober 2024, lalu diproses melalui pembersihan data, normalisasi, tokenisasi, dan pembobotan menggunakan Term Frequency-Inverse Document Frequency (TF-IDF). Model klasifikasi menghasilkan akurasi 78,33%, Presisi 79,25%, Recall 78,33%, dan F1-Score 78,49% dengan sentimen positif dominan, mencerminkan dukungan penggemar yang kuat, meski sentimen negatif dan netral juga muncul. Penelitian ini memberikan wawasan tentang reaksi penggemar dan membuktikan efektivitas analisis sentimen dalam memahami interaksi di media sosial. Abstract. BTS is currently on hiatus as some members fulfill their mandatory military service in South Korea. Although they are not active as a group, the individual achievements and collaborations of the members continue to attract attention. However, negative issues circulating on social media have the potential to affect the public's perception of this group. This study aims to analyze Twitter users' sentiment towards BTS during their hiatus using the Naïve Bayes algorithm, which is effective for text sentiment analysis. Data was collected using crawling techniques on tweets related to BTS during May–October 2024, then processed through data cleaning, normalization, tokenization, and weighting using Term Frequency-Inverse Document Frequency (TF-IDF). The classification model produced an accuracy of 78.33%, Precision of 79.25%, Recall of 78.33%, and an F1-Score of 78.49% with a dominant positive sentiment, reflecting strong fan support, although negative and neutral sentiments also appeared. This research provides insights into fan reactions and demonstrates the effectiveness of sentiment analysis in understanding interactions on social media. ]]>
