<![CDATA[Abstrak - Perkembangan layanan financial technology (Fintech) di Indonesia telah mengubah pola transaksi dan inklusi keuangan masyarakat. Namun, respons pengguna terhadap layanan ini beragam, sehingga diperlukan analisis sentimen untuk memahami opini publik secara objektif. Penelitian ini bertujuan untuk menganalisis sentimen pengguna Twitter terhadap layanan Fintech di Indonesia dengan menggunakan algoritma Naive Bayes. Data diperoleh dari tweet yang mengandung kata kunci terkait Fintech seperti "Fintech", "dompet digital", "pinjaman online", dan "QRIS" dalam rentang waktu Januari–Desember 2023. Metode penelitian meliputi: (1) pengumpulan data menggunakan Twitter API, (2) preprocessing data (cleaning, case folding, tokenizing, stopword removal, stemming), (3) ekstraksi fitur TF-IDF, (4) klasifikasi sentimen menggunakan Naive Bayes, dan (5) evaluasi model dengan metrik akurasi, presisi, recall, dan F1-score. Hasil penelitian menunjukkan bahwa algoritma Naive Bayes mampu mengklasifikasikan sentimen dengan akurasi sebesar 85,2%, dimana mayoritas tweet (62%) memiliki sentimen positif, diikuti oleh netral (28%) dan negatif (10%). Analisis lebih lanjut mengungkap bahwa sentimen positif didominasi oleh diskusi tentang kemudahan transaksi, sementara sentimen negatif terkait keluhan layanan pelanggan dan keamanan data. Temuan ini memberikan implikasi praktis bagi penyedia layanan Fintech untuk meningkatkan kualitas layanan dan mitigasi risiko. Penelitian ini juga membuktikan bahwa Naive Bayes, meskipun sederhana, efektif untuk analisis sentimen media sosial dengan dataset terbatas. Rekomendasi untuk studi selanjutnya mencakup penggunaan algoritma hybrid atau analisis temporal untuk memantau perubahan sentimen secara dinamis.Kata kunci: Analisis; Fintech; Naive Bayes; Twitter; Machine Learning; Abstract - The development of financial technology (Fintech) services in Indonesia has changed the pattern of transactions and financial inclusion. However, user responses to these services vary, so sentiment analysis is needed to understand public opinion objectively. This study aims to analyze the sentiment of Twitter users towards Fintech services in Indonesia using the Naive Bayes algorithm. Data was obtained from tweets containing Fintech-related keywords such as “Fintech”, “digital wallet”, “online loan”, and “QRIS” in the time span of January-December 2023. The research methods include: (1) data collection using Twitter API, (2) data preprocessing (cleaning, case folding, tokenizing, stopword removal, stemming), (3) TF-IDF feature extraction, (4) sentiment classification using Naive Bayes, and (5) model evaluation with accuracy, precision, recall, and F1-score metrics. The results showed that the Naive Bayes algorithm was able to classify sentiment with an accuracy of 85.2%, where the majority of tweets (62%) had positive sentiments, followed by neutral (28%) and negative (10%). Further analysis revealed that the positive sentiment was dominated by discussions about ease of transactions, while the negative sentiment was related to customer service complaints and data security. The findings provide practical implications for Fintech service providers to improve service quality and risk mitigation. This study also proves that Naive Bayes, although simple, is effective for social media sentiment analysis with limited datasets. Recommendations for future studies include the use of hybrid algorithms or temporal analysis to monitor sentiment changes dynamically.Keywords: Analysis; Fintech; Naive Bayes; Twitter; Machine Learning;]]>
