By Mya Kidson
University of Guelph research suggests that Twitter’s big-data capabilities make it useful for improving disease surveillance and public safety.
Dr. Rozita Dara, a computer science professor in the College of Engineering and Physical Sciences, and Dr. Shayan Sharif, a pathobiology professor at the Ontario Veterinary College, have created a system that combines machine learning and artificial intelligence (AI) to detect and track the spread of disease outbreaks. The research team has used this system to monitor COVID-19.
“While vaccination and rapid testing remains the best way to curb the spread of the virus, early warnings can provide a vital solution to predict and mitigate outbreaks from occurring,” says Dara.
Researchers are using social media platforms, such as Twitter, and Google searches for their study.
People tweet about anything, especially hot topics such as COVID-19, says Dara. These tweets could include outbreaks in a certain community or about someone sharing their positive test results.
AI can filter through many tweets about COVID-19, weeding out unrelated posts to collect and analyze relevant tweets that determine the severity and location of outbreaks. This can potentially suggest optimal action plans to mitigate the disease before it spreads to different areas.
This research was adapted from crawler software developed by former computer science PhD student Dr. Samira Yousefinaghani. The software was originally used to identify and monitor Twitter posts about avian influenza, a viral infection that can kill chickens and cause severe economic losses for the poultry industry.
From July 2017 to December 2018, the software searched for and identified 210,000 tweets related to avian influenza from 116,000 unique users.
Monitoring Twitter can help authorities learn about outbreaks and help organizations such as the Public Health Agency of Canada respond promptly to disease spread between neighbouring communities.
This system can be implemented with current pandemic protocols: vaccination, rapid testing, mask-wearing and physical distancing.
“We’ve used this monitoring system for disease modeling, including avian influenza and COVID-19, so there’s no limit to what disease we could track,” says Dara. “Using social media to prevent, track and mitigate the spread of infectious diseases has the long-term potential to be an effective disease mitigation strategy for many more future outbreaks that may arise. Analyzing social media can also capture public concerns at the time of pandemic and enable public health agencies to adopt strategies to manage misinformation.”
The research on avian influenza is funded in part by Food From Thought and the Ontario Agri-Food Innovation Alliance, a collaboration between the Ontario Ministry of Agriculture, Food and Rural Affairs and the University of Guelph.