,
15 tweets,
11 min read
Read on Twitter
[Thread] Interesting phenomenon brought to my attention by @karamballes regarding pro-Brexit accounts trying to rally followers on Twitter today. Firstly if you type "need to increase my pro-Brexit followers" on Twitter you'll see a lot of similar tweets #StopTheCoup 
@karamballes So I examined the accounts and the retweets of any of the accounts that mentioned the phrase "need to increase my pro-Brexit followers". I then did some basic corpus linguistics analysis on those accounts, and again I found a few things quite bizarre.
@karamballes Understandably, #Brexit is the top word featuring in most biographies of the accounts. It appears in around 603 of the 3256 accounts. That's a lot, and could suggest such a call to arms really gets the issue-based Twitter accounts together. 145 accounts contain the word Trump
@karamballes What I found weird, is not that #Brexiteers like Trump, but that Trump is an American politician, and his name appears far more than any British politician. In fact, around seven times more than the next highest politicians.That is Boris & Tommy Robinson, who only appear 23 times
@karamballes What made me laugh even more though, is that @Nigel_Farage 's name only appears 9 times (sorry Nige!). What's even funnier, is that Corbyn's names appears more than Farage - 18 times. Yeah it's all anti-Corbyn bile, but that's totally on message right?
@karamballes So what we can observe from this Brexit twitter surge, in terms of pure numbers, is that Trump > Islam > Boris/Tommy> Corbyn > Farage. Oh yeah, I stuck Islam in there cos it's mentioned 30 times, almost exclusively in a negative context. Maybe they want @KTHopkins to follow them?
@karamballes @Nigel_Farage The phrase "anti EU" appears 56 times (out of around 3256 accounts remember), and the phrase 'anti-' is very common. Appearing 137 times. "Hate the EU" appears 25 times, while "hate" appears 124 times. Indeed, the hate and anti is strong...But not as bad as the love!
@karamballes @Nigel_Farage Just so you don't think these guys are a bunch of negative nellies, the word love appears 394 times. What do they love, well 65 LOVE EUROPE (but HATE or DETEST the EU). Where have I heard that one before?
@karamballes @Nigel_Farage A weird amount say they love their family, dogs, and animals...Despite there being a lot of Islamophobes, the Brexiteer-Israel nexus is strong, with 21 people confessing their general support and/or love for Israel. (That last one may seem odd, but I see it a lot with bots)
@karamballes @Nigel_Farage Anomaly detection is a little difficult (i.e. spotting suspicious spikes in account creation) in a sample this size, but there are a few spikes above the standard deviations (May 2011 for example). What does that mean - probability of inorganic accounts of course...
@karamballes @Nigel_Farage So what can we deduce from this pro-Brexit Twitter storm. Firstly,are the accounts real? It's quite hard to say. However, let's assume they are real first. This would mean that they are more likely to be mentioning being pro-Trump than any other British politician added together!
@karamballes @Nigel_Farage If the accounts are real, this might suggest Trump's interventions on Brexit matter a lot, because clearly Trump matters a lot to this group. Of course if many of the accounts are fake, which is highly probable given the unusual Twitter behaviour, and the seeming lack
@karamballes @Nigel_Farage of biographical diversity, then we have to ask ourselves again why a network predominantly consisting of pro-Trump accounts is tweeting on #Brexit? Also, why is something like Israel so relevant to this group? Is it the global right standing in solidarity, or not? I am curious as
@karamballes @Nigel_Farage to why there is the drive for accounts to get more followers. Beyond ego, I imagine it has to do with making suspicious accounts look more legitimate by having more followers and interconnected relationship on Twitter. This in theory would make them less likely to be flagged
@karamballes @Nigel_Farage Next step would be to determine how likely it is for such terms to appear in relation to a defined corpus. If certain words are used so frequently it could imply an organised Twitter PR campaign, or simply that the level of debate, polarisation and hate in the UK is at its nadir
