KAIST system uses AI to predict inbound Covid patients

Home > Business > Industry

print dictionary print

KAIST system uses AI to predict inbound Covid patients

A diagram demonstrates the Hi-CovidNet system that can predict the number of inbound Covid-19 patients to Korea. [KAIST]

A diagram demonstrates the Hi-CovidNet system that can predict the number of inbound Covid-19 patients to Korea. [KAIST]

With the number of Covid-19 infections surpassing 20 million globally, Korean researchers have developed a system designed to predict the number of virus patients entering Korea from abroad.
A KAIST research team led by Lee Jae-gil, a professor teaching industrial and systems engineering at the Daejeon-based university, announced on Wednesday that the research team has developed a system that is able to predict the number of infected patients likely to enter Korea over a two-week period. Named Hi-CovidNet, the system is based on big data and artificial intelligence (AI).
The system uses AI to analyze relevant big data, such as the number of web searches related to Covid-19, scheduled flights and people who sign up for roaming services for Korea from overseas. Information on the number of people entering Korea in real time wasn’t used due to data restrictions.
Making the prediction could help the government expand quarantine facilities, secure a sufficient number of test kits and ensure enough people are working at quarantine sites. The information could also be utilized to help the government set up quarantine policies against certain countries, according to the research team.
Precisely distinguishing the channel through which Covid-19 has spread in Korea is important to minimize the effect of the virus. Infections through inbound patients are one of the crucial factors that contribute to the spread of the virus, according to the research team.
The research team also highlighted the role of the geographical distance between inbound and outbound countries in affecting their predictions. 
“Covid-19 is more easily spread to neighboring countries, and [the spread of infection] is also affected by the level of exchange between the countries,” said the research team, adding that it set up the technology to learn the connections between countries based on their geographical proximity.  
According to the research team, Hi-CovidNet is up to 35 percent more accurate in predicting the number of inbound Covid-19 patients than the existing model.
“The research is an example that shows how the latest AI technology can be applied in preventing Covid-19,” said Kim Min-seok from the research team. Hi-CovidNet “is expected to raise Korea’s reputation in preventing the pandemic.”
KT and the Ministry of Science and ICT supported the research by providing information on roaming services.
BY JIN MIN-JI   [jin.minji@joongang.co.kr]

Related Article

More in Industry

Majestic flight

Hyundai Motor is new darling of the stock market

Doosan Tower sells for 800 billion won as financial woes continue

Hyundai E&C wins big rail project contract in Philippines

Spud sundaes and ugly apples as retailers rush to help farmers

Log in to Twitter or Facebook account to connect
with the Korea JoongAng Daily
help-image Social comment?
lock icon

To write comments, please log in to one of the accounts.

Standards Board Policy (0/250자)

What’s Popular Now