Predicting Depression Levels Using Social Media Posts

Abstract

The use of Social Network Sites (SNS) is increasing nowadays especially by the younger generations. The availability of SNS allows users to express their interests, feelings and share daily routine. Many researchers prove that using user-generated content (UGC) in a correct way may help determine people's mental health levels. Mining the UGC could help to predict the mental health levels and depression. Depression is a serious medical illness, which interferes most with the ability to work, study, eat, sleep and having fun. However, from the user profile in SNS, we can collect all the information that relates to person's mood, and negativism. In this research, our aim is to investigate how SNS user's posts can help classify users according to mental health levels. We propose a system that uses SNS as a source of data and screening tool to classify the user using artificial intelligence according to the UGC on SNS. We created a model that classify the UGC using two different classifiers: Support Vector Machine (SVM), and Naïve Bayes.