Early Detection of Depression Indication from Social Media Analysis

¹ D.J .Sanghvi College of Engineering, Mumbai.
² Department of Computer Engineering, D.J.Sanghvi College of Engineering, Mumbai.

^* Corresponding author: mahrukhsyed11@gmail.com
meera.narvekar@djsce.ac.in

Abstract

Depression that stems through social media has been steadily growing since the past few years but with the current inclination towards social media reliance it is highly imperative to detect the early signs. Continuous observation of a user's social media interests and activities may highlight suspicious and negative thoughts. This observation can help in understanding their future course of action and also indicate any suicidal thoughts and behaviors. By using the machine learning models, early indications of depression detection can be addressed. This work studies different word embedding techniques for early detection of depression from social media posts. Further, this work develops a model using various NLP processes in order to address the issue of early detection. The recommendations can be useful as a Decision Support System for counselors, psychologist and also can be of good use by the cyber-crime cell department for criminal investigations.