The story of how a Canadian company warned the world of coronavirus nine days before WHO did.
Late last year, COVID-19 (widely known as coronavirus) surfaced in a seafood market in China. Dozens of cases with leading symptoms of pneumonia and fever were being treated, the government of Wuhan confirmed on December 31. As COVID-19 has never been seen before, the only sources of information that were available to predict the spread of the virus at that time were social media, reports… and airline data.
After seeing the SARS virus firsthand in Toronto, Dr. Kamran Khan, a physician specialized in infectious diseases, was looking for a better way to anticipate an outbreak. From his efforts, Bluedot, a Toronto-based company was born. Bluedot provides predictions for governments and health organizations of upcoming epidemics using artificial intelligence.
They gather news and reports from hundreds of thousands of sources in 65 languages. For this vast amount of articles, they utilized natural language processing and machine learning techniques to extract and categorize reports. This is how they first came across the escalating situation in Wuhan. Given an unknown, possibly infectious virus, what is the fastest way for a virus to spread? Airlines, of course. With that in hand, they correctly predicted Bangkok, Hong Kong, Tokyo, Taipei, Phuket, Seoul, and Singapore to be the next major cities infected.
Other than news, reports, and flight data, they also utilize livestock health reports and population demographics. Though the technique proved to be successful in predicting the Zika virus in 2016 and Ebola in 2014, it does not replace human expertise, as Khan explained to CNBC:
We don’t use artificial intelligence to replace human intelligence, we basically use it to find the needles in the haystack and present them to our team,
While predictions themselves do not stop epidemics, it can give invaluable insights to the already overloaded governments and health organizations on how to organize their resources.