Using AI To Track Changes In Sentiment Around Climate Change

There are numerous seemingly contentious topics at the moment, with climate change and vaccinations high among them.  Research from the University of Waterloo aimed to use AI to determine whether our opinions on these hot-button topics have changed over the years.

After analyzing tweets from 2007 to 2016, the researchers found that the debate around climate change had not really shifted a great deal, but that wasn’t the case with regards to vaccines.

Changing sentiment

For climate change, this means that sentiment that was overwhelmingly in the believe camp remained there.  These people strongly believe that climate change is caused by human activity, and is, therefore, something that requires concerted and coordinated action.

Vaccine sentiment was very different, however, with only around 15% of users expressing a clear pro-vaccine stance and the overwhelming majority largely ambivalent on the topic.  What perhaps was most telling in relation to the subsequent debate during the pandemic is that there seemed to be much less discussion among those with different vaccine stances than those with different stances on climate change.

“It is an open question whether these differences in user sentiment and social media echo chambers concerning vaccines created the conditions for highly polarized vaccine sentiment when the COVID-19 vaccines began to roll out,” the researchers say. “If we were to do the same study today with data from the past two years, the results might be wildly different. Vaccination is a much hotter topic right now and appears to be much more polarized given the ongoing pandemic.”

Finding connections

The researchers were especially keen to see if there were any connections in how hot-button topics are discussed, whether in terms of who is discussing them or how they find and share information.

“There’s been some work done on the polarization of opinions in Twitter and other social media,” they explain. “Most other research looks at these as isolated issues, but we wanted to look at these two issues of climate change and vaccination side-by-side. Both issues have social and environmental components, and there are lots to learn in this research pairing.”

The researchers analyzed around 87 million tweets in total, ranging from 2007 to 2016.  AI was then used to rank each tweet according to whether it was pro, anti, or neutral on the given topic, with users classified similarly.  The system also aimed to analyze the structure of the online communities, including the degree of interaction between users from different communities.

“We expected to find that user sentiment and how users formed networks and communities to be more or less the same for both issues,” the researchers conclude. “But actually, we found that the way climate change discourse and vaccine discourse worked on Twitter were quite different.”

Facebooktwitterredditpinterestlinkedinmail