ANALISIS SENTIMEN OPINI MASYARAKAT TERKAIT PENGGUNAAN MOBILE JKN PADA MEDIA SOSIAL X MENGGUNAKAN ALGORITMA K-NEARESTNEIGHBOR (KNN) DAN LOGISTIC REGRESSION (LR)

Hematang, Kelvin

Jurnal Sistem Informasi
Vol 6 No 2 (2025): Jurnal Sistem Informasi

BPJS Kesehatan launched the Mobile JKN application in 2017 to enhance healthcare accessibility and simplify membership management. As of September 2024, 98.67% of Indonesia’s population has joined the JKN program, highlighting the importance of public opinion in the application’s development. This study analyzes public sentiment toward Mobile JKN through the social media platform X using the K-Nearest Neighbor (KNN) and Logistic Regression (LR) algorithms. From 4,777 reviews, 4,648 clean data were obtained after preprocessing and labeled using IndoBERT, resulting in 2,066 positive reviews and 2,581 negative reviews. The evaluation results show that KNN achieved an accuracy of 83.87%, while LR performed better with 87.85% along with higher precision, recall, and f1-score. The findings reveal the dominance of negative sentiment in user reviews and provide insights for BPJS Kesehatan to improve Mobile JKN into a more responsive application that meets users’ needs and feedback.