We believe our project can be used to predict hotspots of disease spread and help local governments find the best areas to direct resources.
Using socio-economic information from Census Tracts (data such as poverty rate, education, race-ethnicity, population pyramid, proximity to health care, and old age, as well as NASA population density data, we build heatmaps to show how at risk communities are through spatial data analytics and machine learning. Data is extracted and stored in a spatial data model, and we use machine learning to evaluate the risk factor based on published statistics as weights. We display our risk heatmap on a web application through the ArcGIS javascript API.
In addition, in our web app, we built a chat AI to answer any questions the user may have, and built a web scraper to scrape reliable local news.
Our web app is fully integrated with Google Cloud, running on App Engine, and using Places API and Dialogflow for the chat AI. We used React.js/Node.js for our web stack and conducted scraping with Node.js using Puppeteer.
NASA SEDAC Gridded Population of the World, Population Data v4.11, 2020.
US Census Data, Census Bureau
APM Research Laboratory Publications
World Health Organization Publications