Understanding the user’s need for health information has become increasingly important as the use of digital health services continues to grow. However, the unstructured data of user-generated questions presents challenges in accurately capturing and analyzing these needs. This study contributes to addressing SDG 3 (Good Health and Well-being) by utilizing topic mining-based knowledge discovery to identify the primary topics emerging from user questions submitted through the “Tanya Dokter” feature on the Alodokter platform. A total of 8, 550 questions were obtained through web scraping between July 2024 and June 2025. The collected data were preprocessed and subsequently analyzed using seven topic modeling approaches: Latent Dirichlet Allocation (LDA), Correlated Topic Model (CTM), Latent Semantic Analysis (LSA), Non-negative Matrix Factorization (NMF), BERTopic, Top2Vec, and ProdLDA. To assess model performance, the coherence metric (cᵥ) was employed to identify the most effective method. Among these techniques, NMF achieved the best results, producing the highest coherence score of 0. 67 with six well-defined topics. The findings show six primary areas of concern: pregnancy; menstruation and contraceptive management; general health and minor ailments; infant care; dermatological conditions; and musculoskeletal and other physical complaints. General health-related issues occurred most frequently, particularly during seasonal transitions, while menstruation and contraceptive management received the least attention, despite menstruation contributing to women’s health risks and the use of contraceptives helping to reduce maternal mortality in Indonesia. These findings offer valuable insights for digital health platforms like Alodokter to enhance information delivery and health literacy, ultimately improving online health services and supporting the achievement of SDG 3
Harahap et al. (Mon,) studied this question.