Detecting Depression in Social Media Posts Using Machine Learning

Abstract

The utilization of Social Networking Sites (SNS) like Twitter is expanding quickly and particularly by the more youthful age. The profit capacity of SNS enables us to express their interests, emotions and offer their day by day schedule. SNS sites such as Twitter allow for constant investigation of user behaviour. Such examples are important for the psychological research network to comprehend the periods and area of most prominent interest. Worlds fourth biggest disease depression has turned out to be a standout amongst the most huge research subject. We propose a system which uses tweets as source of data and SentiStrength sentiment analysis to create a training data for our system and a Back Propagation Neural Network (BPNN) model to classify the given tweets into depressed or not depressed categories.