Media Releases

While the long-term emotional impacts of the COVID-19 pandemic are not yet fully understood, social media is providing valuable insight into how isolation may be affecting our emotional well-being.

Researchers at Dalhousie University are using AI to analyze millions of tweets to identify trends that could help inform public health policy and guide potential early interventions to support people’s mental and emotional well-being.

The team from Computer Science and Community Health and Epidemiology looked at real-time public reactions using the hashtag #stayathome to understand how those in a 14-day quarantine were coping emotionally with the precautionary measure.

Dr. Rita Orji, Canada Research Chair in Persuasive Technology and associate professor in the Faculty of Computer Science, used a deep learning model to classify tweets in eight prominent emotion categories, such as anticipation, anger and happiness. Dr. Orji conducted the work with Dr. Stan Matwin, Canada Research Chair and director of the Institute for Big Data Analytics, and Dr. Swarna Weerasinghe, associate professor in the Department of Community Health and Epidemiology.

The team could then identify trends identifying the point in the isolation at which most people felt certain emotions, such as anxiety. The research suggests that anger was one of the most common emotions, with people growing increasingly angry over time.

Dr. Orji is available to discuss the findings and how machine learning can help predict where people are emotionally at a point in time and where they might be over the course of their isolation -- information that could be extremely useful for public health experts when developing strategies to help people at times like this.

Media contact:

Alison Auld
Senior Research Reporter
Dalhousie University
Cell: 902-220-0491
Email: Alison.auld@dal.ca