Twitters users less likely to support Brexit, say Irish data scientists
Researchers at the Insight Centre for Data Analytics have analysed 1.5m tweets mentioning the UK’s referendum on Europe over the last two weeks. As of Tuesday morning (June 21st) 62 per cent of tweets expressed support for the UK staying in the EU. Britons will vote in a referendum tomorrow on whether to stay in or leave the EU.
The SSIX consortium is a small group of data researchers and tech companies focused on data analytics. The group measured English-language tweets about Brexit from around Europe, counting 73.1 per cent of activity from Twitter users in the UK.
After the UK, Twitter users in the Netherlands issued the most Brexit tweets, more than 10 per cent of all the tweets measured.
Tweets from Ireland made up around 3 per cent of the total, while Greek Twitter users accounted for 2 per cent of tweets.
Belgium, France, Germany, Spain and Italy each made up less than 2 per cent of the total number of tweets on Brexit.
Data researchers started mining tweets on 10 June, measuring whether they expressed strong or weak support for Brexit, or strong or weak support for remaining in the EU.
SSIX partnered with European news agency EurActiv.com and shared the data it aggregated on Brexit. German newspaper Handelsblatt is also a partner in the consortium.
British humor made it hard to judge whether some Twitter users were for or against Brexit, researchers revealed.
“The irony, especially the unique British dry sense of humour, was a challenge,” said Dr Brian Davis of the Insight Centre for Data Analytics, who is coordinator and leader of the SSIX Consortium . “For example, one tweet read: ‘Everybody tell all their #Brexit colleagues at work that the referendum is June 24 – they’ll believe anything’.
“It’s not an explicit stay or leave but an implicit criticism of supporters of leaving the EU and this is extremely hard to categorize.”
Researchers used anonymous data and set up an algorithm to categorise tweets automatically. They didn’t record Twitter profiles or whether tweets were retweeted, liked or went viral.
