Are we what we watch: Analyzing the effect of YouTube videos using comments on viewer’s mental health using Python
Shanya Sharma (~shanya) |
Ever wondered what that dark YouTube video you watched did to your mind? While YouTube has become one of the most used applications in today's era, we can't deny the ill effects it has on us. Not just that, the addiction of watching the wrong kind of videos can lead our mind in various troubles. Depression being one of them.
Depression, a serious mental health issue, is affecting a significant segment of the society today. The estimated prevalence of depression is as high as 300 million people around the world, according to the World Health Organization, which proves to be a major reason for suicide.
In this talk, we'll focus on studying the effect of YouTube videos on the mental state of the individuals watching it by analyzing their comments using python. The comments will be analyzed by extracting their textual features and calculating the sentiment (positive/negative) to understand the connotation of the comment and mapping it to a score called CESD score used by clinics for the screening of depression. We'll also understand what is Empath and learn how to use Empath using python for the generation and extraction of various categories from the text.
- Need for analyzing YouTube Comments [1-2 minutes]
- How do Doctors analyze the severity of depression[2 minutes]
- Current analysis methods and the problems with them[2 minutes]
- Our attempt at fixing it: [14-17 minutes]
Data description and collection methodology
Empath and how it helped (with a DIY)
The Algorithm (Mapping to CES-D score)
- Results [3-4 minutes]
- Using it as an evaluation metric
- DIY (Running the code)[2-3 minutes]
- Q/A - [2 minutes]
Basic Understanding of Python
Interest in analyzing text for depressive triggers
Interested audience can have a look at the actual project and the progress till date here. Links specific to the talk are given below:
Shanya is currently a Software Engineer at SAP Labs India working on building intelligent solutions for Quality Assurance. She is a Machine Learning Enthusiast and has been working on Machine and Deep Learning for almost a year. She has also been a constant admirer of ML/DL advancements and frameworks.
Because of her inclination towards using technology for social good, she makes sure that she's contributing her share by being an active volunteer at platforms like Data4Democracy and DataKind.