Understanding Health Application Adoption: A Classification Approach Using BiLSTM on Individual Factors

In the rapidly advancing digital era, the use of health applications in Indonesia continues to rise, yet the adoption of these applications still faces various challenges, particularly related to individual factors. This study aims to identify and classify the individual factors that influence the adoption of health applications in Kabupaten Bandung. Data was collected through interviews and then analyzed using the deep learning algorithm Bidirectional Long Short-Term Memory (BiLSTM). The study applied two BiLSTM scenarios: the first scenario using a single-layer BiLSTM with a batch size of 16, and the second scenario using a multi-layer BiLSTM with the same batch size. The results showed that the first scenario, a single-layer BiLSTM with a batch size of 16, achieved the highest accuracy of 75.80%. Factors such as Behavioural Intention, Trust, and Self-Efficacy were found to have a significant positive impact on users' decisions to adopt health applications, while Technostress and Habit had a negative impact. These findings highlight the importance of understanding psychological and behavioral factors in promoting the adoption of health applications, which can lead to increased acceptance and usage of these applications among the community.