A Comparative Study of Predicting Mental Distress in Patients via Social Media Interactions: Artificial Intelligence versus Ensemble Machine Learning Approaches

Mental health of a person is a reflection of their emotional, psychological, and social well-being. It establishes one's feelings, thoughts, and reactions to circumstances. With sound mental health, one may reach their full potential and do tasks effectively. But nowadays people are reluctant to discuss their mental issues which makes it very challenging to diagnose mental health. In addition to lowering a person's quality of life, mental illness also increases medical expenses. The potential of AI in mental health is becoming increasingly apparent in that cases. The only way through which a person's mental health can be diagnosed without discussing it is by analysing their behaviour specifically analysing the interactions on social media platforms. In the current scenario, people are proactively involved on many social media platforms on daily basis and hence social media platforms are used to model mental health and evaluate health results accordingly. To foresee the occurrence of specific mental illnesses including anxiety, depression, and suicidal, Researchers are currently applying quantitative methodologies. When these applications are used in real-time, they can help society by acting as a surveillance tool for people who exhibit abnormal behaviour. The focus in this paper is to make a comparative analysis of how different ensemble machine algorithms and methodologies are used to study the behaviour through social media and hence by determining their performance. An effective system for early major depressive system diagnosis can be created with the help of advanced natural language processing (NLP), deep machine learning, and transfer learning techniques.