Social media mental health analysis framework through applied computational approaches

Studies have shown that mental illness burdens not only public health and productivity but also established market economies throughout the world. However, mental disorders are difficult to diagnose and monitor through traditional methods, which heavily rely on interviews, questionnaires and surveys, resulting in high under-diagnosis and under-treatment rates. The increasing use of online social media, such as Facebook and Twitter, is now a common part of people’s everyday life. The continuous and real-time user-generated content often reflects feelings, opinions, social status and behaviours of individuals, creating an unprecedented wealth of person-specific information. With advances in data science, social media has already been increasingly employed in population health monitoring and more recently mental health applications to understand mental disorders as well as to develop online screening and intervention tools. However, existing research efforts are still in their infancy, primarily aimed at highlighting the potential of employing social media in mental health research. The majority of work is developed on ad hoc datasets and lacks a systematic research pipeline. [Continues.]