How One Company Is Using A.I. to Predict Panic Over Covid-19
How One Company Is Using A.I. to Predict Panic Over Covid-19. The Coronavirus Panic Index applies artificial intelligence to human behavior in real-time.
How One Company Is Using A.I. to Predict Panic Over Covid-19, From empty grocery store shelves to empty city streets, there are many visible signs of increased anxiety over the coronavirus crisis. But how much is this concern really permeating people’s lives?
Dayton, Ohio-based behavioral analytics company Cognovi Labs has found a way to measure people’s emotional response toward Covid-19–in real time–with its Coronavirus Panic Index.
The index combines technology–developed at the Wright State University KNO.E.SIS Innovation Center in Dayton, Ohio–with behavioral psychology to analyze conversations across social platforms like Twitter, Reddit, and blogs. The text, and its context, are categorized into emotions (joy, anger, disgust, fear, sadness, surprise) using machine learning. Then it’s all translated into behavioral signals, which are calibrated on a scale of zero to 100–zero being “at home emotionally” and 100 being somewhere near “nervous breakdown.” The points are then plotted on a map.
“It’s not just what people say or how they talk, it’s the emotional sentiment behind that,” says Ben-Ami Gradwohl, co-founder and CEO of Cognovi Labs.
A screenshot taken on March 26 of the Coronavirus Panic IndexCOURTESY COMPANY VIA SCREENSHOT
For businesses, the tool can be used to gauge people’s emotional reactions toward the virus, employment, the stock market, or certain industries or activities, such as shopping or travel.
“We want to understand how emotions translate to action,” Gradwohl says. “We have a chief psychology officer as part of the team to translate that–what people say, how they say it–and, most important, how they felt when they said it, and what that’s going to mean for the future.”
So where are we emotionally as a country? Gradwohl says there was a huge spike in emotional sentiment, especially fear toward the coronavirus early on, at the end of January into February, weeks before the stock market crash–and even before the majority of people in the U.S. were affected. Now, he says, the level of panic overall has started to flatten.
