Uncovering insights in mMedical apps: a text analytics and topic modeling approach

Mobile Medical applications offer a diverse range of services, including medication management, teleconsultation, symptom tracking, and medical monitoring, thereby addressing the needs of both individual users and healthcare providers. This study proposes a systematic methodology for extracting user requirements by categorising medical mobile applications based on descriptions obtained from relevant platforms. The Android platform was selected as the primary data source due to its extensive repository of applications. A total of 976 applications were identified and included in the analysis. To uncover underlying themes, topic modelling was applied to the app descriptions. For each identified topic, the corresponding applications were mapped, and relevant metadata, such as download statistics and user ratings, were systematically analysed and reported. Furthermore, for a selected high-demand category, customer reviews were collected and analysed to extract user requirements. Sentiment analysis was conducted to explore user perceptions regarding app characteristics and potential areas for improvement. To further investigate, a linear regression model was developed to predict user ratings based on the sentiment. The findings of this study provide valuable insights into the functionality of medical mobile applications, offering implications for both practical applications and academic research in the field of digital healthcare.