Sentiment Analysis of JKN Mobile Application in Health Technology Using Naive Bayes, Gradient Boosting, and AdaBoost

Deni Utama (Universitas Negeri Jakarta)
Rauhil Fahmi (Universitas Negeri Jakarta)
Muhammad Ridho Kurniawan Pratama (Universitas Negeri Jakarta)

Jurnal Pendidikan Vokasional Teknik Elektronika (JVoTE)
Vol. 8 No. 2 (2025): Vol 8 No 2 (2025): JURNAL PENDIDIKAN VOKASIONAL TEKNIK ELEKTRONIKA (JVoTE) Volu

The increasing adoption of mobile health services highlights the need to evaluate user experience through feedback. The JKN Mobile application, developed by BPJS Kesehatan (Indonesia’s Social Security Agency for Health), has accumulated thousands of user reviews on the Google Play Store. These reviews provide an opportunity to gain insights into public sentiment regarding the app. This study conducts sentiment analysis using three machine learning algorithms—Naive Bayes, Gradient Boosting, and AdaBoost—to classify user opinions as positive, negative, or neutral. The reviews were preprocessed, vectorized using TFIDF, and evaluated using standard classification metrics. The results indicate that Naive Bayes achieves the highest classification performance, followed by AdaBoost and Gradient Boosting. These findings offer valuable guidance for developers and policymakers in improving the app’s usability and functionality.