Opinion mining social analysis indicates decline in criticism of the UK’s move to leave the EU
Using a unique opinion mining approach to data analytics around UK social media conversations between 14th and 29th March 2017 it was revealed that a relatively small proportion of the British public discussing Brexit now showed a negative opinion of Article 50 being triggered.
The data showed that negative sentiment towards Brexit accounted for only 27 per cent of online mentions, which is in stark contrast to the 48 to 52 per cent outcome of the EU referendum. The data also demonstrated that a significant proportion of users had begun to show neither positive nor negative opinion towards the triggering of Article 50, suggesting that emotive reactions to Brexit have cooled since the referendum.
In addition to the drop in criticism of Brexit, negative sentiment towards Theresa May has fallen significantly. BrandsEye recently showed that 55 per cent of online mentions of Theresa May between 20th January and 24th February were negative. These unfavourable perceptions of the Prime Minister centred on her relationship with Donald Trump and talks of his proposed state visit to the United Kingdom.
A large amount of this negative sentiment has since dissipated and after the latest analysis only 12.8 per cent of conversations expressed disapproval of the Prime Minister. It must be noted however that the positive sentiment remained low and neutral conversation increased significantly.
Negative opinion of Theresa May has tended to focus on an opposition to the relationship between the UK Prime Minister and President of the United States, Donald Trump. It will be clearer in the coming weeks how public opinion is affected by the UK’s continuing quarrel with Spain over the status of Gibraltar and upcoming talks with the EU.
This data suggests that a significant proportion of Remain voters have come to terms with the UK’s future. Those on that side of the debate are now much less vocal in their opposition to leaving the EU.
In relation to the Prime Minister, she has recovered from her difficult trip to Washington in the eyes of the public and demonstrated strong leadership in the aftermath of the terror attack in London.”
Overall conversation volumes were 50 per cent greater in the week leading up to 29th March compared with the week before. A total of 1,075,509 tweets mentioned ‘Brexit’ and ‘Article 50’ between 22nd and 28th March. Conversation on the day that Article 50 was triggered – 29th March – reached a peak volume of 991,686 mentions.
As part of the research project all mentions of Article 50 and Theresa May in online conversations between 14th and 29th March 2017 were tracked. The data was then analysed through a combination of Crowdsourcing for increased data accuracy and AI.
Last year, BrandsEye used the same methodology to predict both the election of Donald Trump and Brexit. In both instances, popular sentiment displayed on social media provided an accurate reflection of the voting public’s intentions in the US and the UK.