Ranking Mental Illness among Social Media Users

Currently we are in the era of the internet. We can express our thoughts through social media i.e. Facebook, Twitter, Instagram etc. Now we are in the most challenging stage of this century as well as the rest of the world due to COVID-19 pandemic and the after effects of the virus infection. Some of the survivors are experiencing trauma and other mental health issues. To determine the mental health condition of someone, we performed sentiment analysis on their social media posts. This paper focuses mainly on sentiment analysis of twitter data written in Bengali language and tries to find out whether the user is mentally depressed or not. The expressions of those users are either negative or positive, sometimes neutral. Using different contemporary Machine learning algorithms like Naive Bayes, Support Vector Machines (SVM), we can determine the mental health condition of any user. We also built a model by applying the Random Forest classifier on social media posts and examined them whether they were in a mental condition or not.