MACHINE LEARNING TO PREDICTING DEPRESSION LEVEL IN SOCIAL MEDIA DATA LIKE TWEETS

Abstract

One of the most significant distinctions between today's youth and young adults and previous generations is that they spend far less time connecting with people in person and far more time connecting electronically, primarily through social media. The rise in depression, according to some experts, is proof that social media users are less emotionally content. Twitter has become the most popular social media tool for users to communicate information in real time through tiny messages known as tweets. Depression is a prevalent long-term illness. Due to limited detection procedures, it frequently goes unnoticed, resulting in major consequences for public and personal health. Depression is a big worry that is becoming more prevalent by the day. Depression can have a variety of reasons, but the most common is mental disease. Many people suffer from depression, yet only a small percentage of them seek treatment. The goal of this research is to apply machine learning approaches to detect a possible depressed Twitter user's tweet. We used variables collected from a user's behaviours within tweets to train and test classifiers to distinguish whether a person is depressed or not. On a scale of 0-100 percent, classification machine algorithms are used to train and classify it in different stages of depression. Also, data was collected in the form of tweets, which were categorised into whether the person who tweeted was depressed or not using Machine Learning classification algorithms. Early detection of depression or other mental diseases using a predictive technique.