Case Western Creates Coronavirus Risk Assessment Tool

(TNS) — Case Western Reserve University researchers are working to meld public information and user data to track the risk of contracting the coronavirus in any location.

In less than two months, a team of two professors and six students created an online risk-assessment tool with the goal of helping people choose where to go to a park or a pharmacy – across town, or across the nation.

The tool uses artificial intelligence to analyze data from the U.S. Census, state and local health departments, Google traffic maps and social media posts, as well as a risk perception rating submitted by users for any specific spot.

The perception rating is relative, says Professor Yanfang (Fanny) Ye. The aim is to compare specific locations within a region.

So, because of the weighting of the perception factor, for example, you can’t use the risk assessment to compare big cities, like Cleveland to New York. (On Thursday afternoon, Cleveland had a higher risk score than New York City.)

The tool is in its beta stage. And while professors Ye and Kenneth Loparo admit it has challenges, they say the assessment will become more useful as more data sources are added.

“That data might be useful for somebody in the context of everything else you have,” Loparo said. “If might not be the optimal computation of the risk index because of the data available... But as that data becomes more granular, it will allow us to factor in more information that is very, very specific to region.”

You can find the tool here: https://covid-19.yes-lab.org/. You can use the website on a computer – just make sure you allow your browser to give the site location access. Or you can use it on your smartphone.

The site analyzes data, including:

• disease data from public health organizations, such as state health departments and county health boards
• U.S. Census demographic data, including population, businesses, and geographic statistics for all states and counties, and for cities and towns with more than 5,000 people
• traffic data from Google maps, to estimate how busy the area is
• posts from discussion threads on Reditt

Some data, such as the census information, is stagnant. Other information, such as the number of cases in a county or ZIP code, is updated whenever the public health officials release it. But traffic data is updated in real time, as is the risk perception. Users click a rating, from extremely low to extremely high.

The end result is a risk assessment between 0 and 1.

Since the beta version launched Monday, the site has had 40,000 visitors, Ye said.

“We are starting to get feedback from the user,” she said. “We keep improving our system. This will provide you another piece of information you can reference.”

©2020 The Plain Dealer, Cleveland, Distributed by Tribune Content Agency, LLC.