Natural Language Processing COVID Attitudes Tracker

'COVID fatigue' becomes a greater problem the longer the pandemic wears on. People are tired of being told to stay at home and social distance, and the fear factor has lost its edge, leading to decreased COVID compliance and increased spread.
So we decided to map these changing attitudes.

what it does

It crawls Twitter for current tweets, uses deep learning NLP to classify them, and builds a cute little map to visualize our data

how it does it

This bad boy uses Tweepy to acquire tweets, AllenNLP + Pytorch for model training and deployment, Flask for backend, and CSS/HTML/Bootstrap/JS therein to deploy this website you're reading.

why it's relevant and

where we're going from here

We're all so sick if quarantining, so why not get it over with? This technology is not only useful to recognize where COVID compliance is low in order to put better measures in place, but could also be expanded to check security footage to see if people are wearing masks by training a new program to recognize masked vs. bare faces. Some may say surveillance state, others say practical public health, but either way it's a fun hack!