Updated: 4:19 p.m., Tuesday, April 21, 2020
Two local researchers have developed a web tool that aims to show real-time COVID-19 risk at different locations.
Case Western Reserve University faculty members Fanny Ye and Ken Loparo have created an online mapping tool called Alpha-Satellite that attempts to show the risk of community spread of the coronavirus in any given area.
“We hope through this AI-driven system, we can help to allocate and provide the signs of how to prevent and also slow down the spread of the virus,” said Ye.
Alongside Loparo and their research team, Ye compiled data over the past month and created a complex algorithm to determine and map COVID-19 risk at locations across the country.
Ye said the goal is to help users who may be going to a grocery store or other essential business decide which location to visit based on the relative risk of getting the virus there.
Factors that go into that relative risk number and ranking include the number of confirmed COVID-19 cases in the area, how congested the location is based on Google Maps estimations, and U.S. Census Bureau Demographic Data.
The higher the number, the greater the relative risk of getting COVID-19.
“There are a lot of technological challenges to overcome,” she said, "especially (with the) situation changing so fast.”
Another factor that goes into the algorithm is public perception of COVID-19 and social distancing rules. Ye and Loparo evaluated social media posts on the discussion board site reddit to determine how well people in certain areas may have been abiding by social distancing.
“If their perception of the risk of the people in that area is low, there’s a greater chance that they’re out and about,” Loparo said. “So that all sort of seems to indicate that that’s a little bit more of a risky place to go than somewhere else might be.”
Loparo added that while this data can be helpful, it should not be the end-all be-all.
“This data exists, but it’s very hard to wrangle with and get it into a place where it’s interpretable and useful, and that’s really what the site is providing, but you need to take it in the context of everything else and adjust your strategies based on the relative risk and all the other data and information you have,” he said.
Ye and Loparo released the site April 1 and are now looking for feedback from public health experts and the general public to improve the data.
Copyright 2020 90.3 WCPN ideastream. To see more, visit 90.3 WCPN ideastream. 9(MDAxMzY2MjQ0MDEyMzcyMDQ5MzBhZWU5NA001))
