How AI Spotted and Tracked the Coronavirus Outbreak 人工智能是如何寻找和跟踪冠状病毒疫情的
John McCormick 约翰·麦科米克
Artificial intelligence is helping epidemiologists identify and track outbreaks faster and more precisely than ever before-including the spread of the new coronavirus that emerged in Wuhan， China.
A small Toronto-based company that tracks infectious disease， BlueDot Inc.， used AI developed in house to send an alert about the coronavirus outbreak the week before major health agencies issued notifications. BlueDot also accurately predicted where the virus would go next.
BlueDot's AI tool， developed in 2018， serves as “an early warning system” that can quickly detect an outbreak by using AI to analyze information from official and nonofficial sources for relevant words and phrases， said Kamran Khan， BlueDot's founder and chief executive.
The BlueDot system gathers information from sources including:
·more than 100，000 news articles from local and international newspapers in 65 languages each day，
·official reports from the Centers for Disease Control and Prevention， World Health Organization， Food and Agriculture Organization and the World Organization for Animal Health，
·real-time climate data from satellites，
·flight information and anonymized flight itinerary data from the International Air Transport Association， a trade group that represents airlines，
·information from government sources， including data on human populations， insects that can transmit infectious diseases， the number of doctors and nurses per capita in an area， and information on economic and political conditions around the world.
The data flows into a central repository that also contains details about more than 150 pathogens.
In the case of the coronavirus， BlueDot's system spotted a number of signals， including references to things like “undiagnosed pneumonia” in Wuhan.
The company's AI system accurately predicted that the virus would spread next to Bangkok， Hong Kong， Tokyo and other cities by analyzing transportation data on where people fly from Wuhan.