Depression Detection from Tweets Along with Clinical Tests

Abstract

Social media has become a massive surge in this generation for everyone to communicate with others, which on one half connects the world while on the other half has a depressing side with people suffering from mental illness. People nowadays express their thoughts and feelings more easily on social media platforms. In addition, numerous studies have proven that by analysing social media posts, we may identify people with mental problems using machine learning. Twitter is one such platform that covers a wide target audience from all parts of the world and from the tweets we can detect the early stage of depression by analyzing its linguistic markers and emotions. Using a sentimental analysis dataset we have created a model with the help of Naïve Bayes Classifier which will support our primary model that will recognize different emotions from the tweets. We've also performed clinical tests by using the MBTI Types (Myer Briggs Type Indicator), a well-known personality test that identifies a person's traits by indicating one of 16 types.