<![CDATA[As Indonesia’s national digital health platform, the SATUSEHAT application is central to the country's healthcare integration. However, maintaining user satisfaction remains a challenge. This study evaluates public sentiment by analyzing over 30,000 Google Play Store reviews using an ensemble approach of machine learning and thematic analysis. To address substantial data imbalance, five base classifiers were compared with three ensemble models using Random Under-Sampling and SMOTE. Results indicate that the Stacking Ensemble with SMOTE outperformed all other models, achieving an Accuracy of 91.1% and an AUC-ROC of 0.919. Sentiment analysis reveals a critical dissatisfaction rate, with 83.3% of reviews being negative. By mapping user complaints to the ISO/IEC 25010 software quality framework, it was identified that functional suitability and reliability account for 80.7% of the reported issues. This research contributes a robust methodology for Indonesian-language sentiment analysis and demonstrates the utility of ISO/IEC 25010 in translating raw user feedback into actionable software engineering requirements. Practically, the findings provide the Ministry of Health with an evidence-based roadmap to resolve critical barriers in authentication, OTP delivery, and certificate retrieval.]]>
Copyrights © 2026
