Inspiration
Current global situation and implications it has on our perception of everyday issues
What it does
Shows scare and cases level based on available data
How We built it
Using NLP and ML techniques, with API created in Python and a browser client
Challenges We ran into
Preprocessing and classification of data (medium accuracy), integration of our modules, size of data
Accomplishments that We're proud of
Created a ML model which is able to classify sentiment of an article based on its TF-IDF vectorization, fast processing on backend side with lots of data, nice style of presentation in browser client, ability to see the site on mobile devices without a problem
What We learned
Some tools for integration and cooperation, new Python libraries for NLP and API, integration of React and OpenStreetMap
What's next for Corona Scare Map App
We can try to develop the app for more countries and their administrative territorities, create better predictive model for article sentiments, add more languages to sentiment prediction (e.g. English), add more data sources
Log in or sign up for Devpost to join the conversation.