<![CDATA[The JKN Mobile application developed by BPJS Kesehatan aims to facilitate digital access to healthcare services for the public. However, user reviews on the Google Play Store show various opinions, ranging from satisfaction to complaints about technical issues. This study aims to analyze user sentiment, evaluate the performance of the Naive Bayes algorithm, and identify the main technical obstacles. The methods used include data scraping from the Google Play Store, text preprocessing (case folding, cleaning, tokenization, normalization, stopword removal, stemming), word weighting using TF-IDF, and classification using Naive Bayes. The results show that negative sentiment dominates (50.7%) compared to positive sentiment (45.9%). The Naive Bayes algorithm achieved an accuracy of 85.61% with an F1-Score of 85.8%, but the low Cross Validation value indicates model instability. The main complaints from users were login and OTP issues (1,632 complaints), complicated interface, as well as errors and bugs. From this analysis, it is concluded that aspects of system stability and verification still require significant improvement. These classification results can serve as a reference for BPJS Kesehatan in prioritizing feature development and technical improvements.]]>
Copyrights © 2026
