Deep Learning Approach for Measuring the Severity of Depression Using Social Media

Depression., one of the world's most common medical illnesses today, has a negative social and financial impact. The emotional computing models concentrate more attention on analyzing specific client postings than chronological conduct patterns, which are essential for spotting changes and improvements in psychological behavior and user behavior. Context, which entails a consistent examination of emotions and traits over time. This article suggests using the TROAD framework to recognize sequential trends over time in depressive user behavior on social media. To recognize sequential patterns from depressive user behavior, the framework determines the optimal interval to analyze each user activity, collects emotional and contextual variables from user input, and models the features into time frames. Today, depression has become a significant issue in our culture. It is a significant contributor to suicide, particularly among teenagers. The nations impacted by the current coronavirus disease outbreak (COVID-19) have suggested lockdown and social seclusion measures. These policies, which lead to interpersonal isolation, have caused depression and mental health issues to become very major issues. Clinical psychologists typically conduct in-person interviews with sad individuals to provide a diagnosis by the clinical depression criteria. However, individuals frequently avoid seeing doctors when they are first experiencing symptoms of depression. Nowadays, more and more people communicate their sentiments on social media.