Liu Jialin: Examining COVID-19 Transmission in Mega Cities in the US and China Using Machine Learning Algorithms

On a team of researchers led by Assistant Professor of Urban Science and Policy, Guan Chenghe, Postdoctoral Fellow of Urban Ecology, Liu Jialin is using sentiment analysis and machine learning algorithms to examine and predict the spread of COVID-19 in multiple high-risks cities around the world. 

The team is examining time series public sentiment dynamics under COVID-19 spread based on social media data such as Twitter and Weibo in the US and China. They will also quantify how regulations and curbs on urban mobility and the impact of climate on COVID-19 transmission in two countries.

^{Liu’s algorithm at work}

Furthermore, these above mentioned modules will be integrated into a novel machine learning algorithm to monitor the real-time Tweets and Weibo posts, in order to predict high-risk events and areas that may be related to the spread of the virus at its early stage. 

The study will expand understanding on how social media outlets reflect and reshape attitudes bearing on COVID-19 outcomes. It will also contribute to and predict how individuals and communities respond to public health emergencies in the future. 

The project is still on-going because the global coronavirus situation is changing rapidly. Liu Jialin deployed an online server at the end of February, which has been collecting Tweets and Weibo posts from the virus centers. 

“In the beginning, there was only one epidemic center, Wuhan,” says Liu. “But recently, we’ve had to focus on 10+ pandemic centers across Asia, the EU, and North America. We are also extending our networks to Africa and India…. The project has already recorded tens of millions of Tweets and Weibo posts”

This project is supervised by Liu Jialin and prepared by undergraduate research assistant Yao Zhirui ’20.