Study uses Twitter to reveal our mental-health state

A Western-led international research team hopes that tweet you just posted – and the millions of others posted to the popular social media site – will provide insights into our collective mental health during the COVID-19 pandemic.

“It’s almost like we’re looking at the footprint of COVID-19 on mental health through social media,” said Daniel Lizotte, a Computer Science and Epidemiology and Biostatistics professor. “This gives us one kind of view of what’s happening, what people are choosing to talk about on social media.”

Lizotte is conducting the new project, Tracking Mental Health During the Coronavirus Pandemic, with colleagues at the University of Alberta and New York University.

Twitter’s daily active user base has grown by 23 per cent during the pandemic, with millions turning to social media to connect and engage with others. This has allowed researchers a view into the experiences of people not available by other means.

Through the use of natural language processing, the team is scouring Twitter for language about COVID-19 related to different mental-health states to get a sense of the different topics that correlate both negative and positive mental-health issues.

“We want to get a more complete picture of what people are experiencing and connect these pieces together when it comes to their mental health,” Lizotte said.

While most conversations express frustration and doubt, there are also positive feelings being shared, including increased family-time activities. With some surgeries being postponed during this time, surgeons have had more time with their families in the past two months than they had in quite some time, Lizotte said.

“We’re wondering about seeing these, as well. People aren’t focussing on the positive,” he added. “There will be negative things, for sure. However, spending more time with family, changes in the pace of work for some, it all plays a role in the discussion, too.”

Lizotte has seen a steady decline of tweets, specifically about COVID, in the past month. It’s a decrease he was not expecting, but makes sense as conversations around the virus have become normalized.

The goal of this study, which runs through October, is to detect the drivers of mental-health issues associated with the pandemic and, in the event of a second wave, prepare and anticipate techniques to better cope.

By including those in health care and health policy in the discussion, the findings of the study could be beneficial when it comes to decision-making during future crises.

“People working in public health can make more aggregate decisions regarding what sort of supports to implement in different communities,” he said, adding the study may soon begin looking at Reddit feeds, as well. “If job loss is really stressing everyone out, then you might do one set of interventions, compared to social isolation concerns and a completely different set of interventions.”

The team plans to share their findings through an online visual analytics and rapid-reporting system, which will identify real-time topics top of mind for social media users.

The work is supported by a CIFAR AI Catalyst Grant, which funds exploration of the COVID-19 pandemic through high-risk, high-reward ideas and projects.

/Public Release. The material in this public release comes from the originating organization and may be of a point-in-time nature, edited for clarity, style and length. View in full here.