Depression and Impaired Mental Health Analysis from Social Media Platforms using Predictive Modelling Techniques

摘要

Depression is the leading global disability, and unipolar (as opposed to bipolar) depression is the 10th leading cause of early death, as stated by the World Health Organization (WHO) in 2015. The study aims to build an approach for depression and impaired mental health analysis from social media platforms. Although for Depression analysis and cure. Psyscologists preferred over machines because they are manipulative and precautionary to Human emotions to a greater extent, Machine Learning has an added advantage. It has no emotions; it studies patterns, not face or beauty or other factors. It studies a wide variety of data and then trains to give better predictions. Although it is not 100% reliable nor are the doctors. Moreover, in countries like India where people don't treat Depression as a Chronic Illness or don't even consider it as an illness of any sort, embedding Machine Learning Depression Detection Algorithms in Social Media combined with recommendation systems to treat a Human Mind positively, still being unnoticeable is a Great Boon to humanity The study is assisted by data collected from users after obtaining their consent and applying data preprocessing techniques. Several machine learning is used to analyze the data in the best way possible. A VAPID Technique is developed that performs far better than a classic feed-forward neural network. This study aims to develop a correlation between features and depressed people to observe a continuous pattern. Moreover, the aim is to conclude that social media can be a new exceptional methodology for analyzing depression and analyzing indirect patterns, improving many lives.