Depression is a common mental disorder involving a depressed mood or loss of pleasure for long periods, which induces grave financial and societal ramifications. Social media-based depression detection is an effective method for early intervention to mitigate those consequences. Such a high-stake decision inherently necessitates interpretability. Although a few studies explain this decision based on the importance of linguistic or demographic features, these explanations do not directly relate to depression diagnosis criteria that are based on symptoms. To fill this gap, we develop a Focused Temporal Prototype Network (FTPNet) to detect depression and provide interpretations based on depressive symptoms as well as their temporal distributions. Extensive evaluations using large-scale datasets show that FTPNet outperforms comprehensive benchmark methods with an F1-score of 0.864. Our result also reveals fine-grained and emerging manifestations of depressive symptoms, such as sharing admiration for a different life, that are unnoted in traditional depression surveys like the Patient Health Questionnaire-9 (PHQ-9). We further conduct a user study to demonstrate improved interpretability over the benchmark. This study contributes to the Information Systems (IS) literature by introducing an interpretable depression detection approach that models the temporal distribution of depressive symptoms. In practice, multiple stakeholders, such as social media platforms and volunteers, can apply our approach to identify depressed users and deliver targeted assistance.

With the growth of online mental healthcare platforms, health professionals have been providing online consultation services during their spare time. However, little is known about the prosocial behaviors of health professionals on these platforms and their effect on professionals’ economic performance. In this research, we aim to identify and quantify the main effect of prosocial behaviors (i.e., offering free services) on future economic performance and the potential mediation effects of relationship capital and reputation capital in an online mental healthcare platform. Based on signaling and commitment theories, we use a panel dataset from a Chinese online mental healthcare platform to test our hypotheses. Our findings show that a mental health professional's previous prosocial behavior has a positive impact on economic performance and that this effect may manifest through the relationship capital pathway. However, reputation capital generated from prosocial behaviors does not significantly mediate the impact of prosocial behavior on economic performance. Our research provides important implications for healthcare operations concerning service offerings and healthcare providers’ performance on online mental healthcare platforms.