New Delhi: Between 60 and 80 per cent of Twitter handles posting on the Russia-Ukraine war may be bot accounts, a research by scholars from the University of Adelaide, Australia has found. Among other influences, these bot accounts may have been pushing people to flee their homes during the conflict between these two countries, the researchers added.
The researchers also found more “pro Ukraine’ accounts than those that were “pro Russia”.
The paper titled “#IStandWithPutin versus #IStandWithUkraine: The interaction of bots and humans in discussion of the Russia/Ukraine war” was released on 20 August.
The researchers studied 5.2 million posts on Twitter — tweets, retweets, quotes and replies to tweets — shared between 23 February and 8 March this year to understand how bot activity may influence online discussions around the Russia-Ukraine conflict and how bots may influence human emotions.
The studies posts contained hashtags like “StandWithPutin”, “(I)StandWithRussia”, “(I)SupportRussia”, “(I)StandWithUkraine”, “(I)StandWithZelenskyy” and “(I)SupportUkraine”.
Bot accounts were identified using Indiana University’s Botometer — a software which helps identify a bot account.
“We can say that between 60 per cent and 80 per cent of accounts tweeting the hashtags we studied during the first two weeks of the war were bots, as determined using the Botometer,” Joshua Watt, one of the researchers, told ThePrint.
According to Watt, it is not clear whether the bots were influencing people to flee Ukraine or Russia.
Watt added: “We cannot conclude where this is happening due to having no geographical information on the origin of accounts. All we can conclude is that the bot accounts are influencing more discussion surrounding moving/fleeing/going or staying in a country/location.”
More ‘pro Ukraine’ bot accounts
According to the researchers, 90.16 per cent of the accounts tweeting on the Russia-Ukraine war were “pro Ukraine” and only 6.80 per cent were “pro Russia”. “Balanced accounts” — those which showed mixed behaviour — comprised 3.04 per cent.
“The ‘ProRussia Not Bot’ account group has the largest outward information flows and significant flows to a range of other groups, having a positive information flow into both ‘ProUkraine’ and “Balanced” account groups, observed the researchers.
This means genuine pro-Russia users have the ability to influence more users on Twitter than genuine users who are pro-Ukraine.
The researchers found “a spike in bots on March 2nd and 4th. The first spike aligns with Russia capturing Kherson (a town in Ukraine), but also when the #(I)StandWithPutin and #(I)StandWithRussia hashtags were trending.”
The research also found noon to 1:00 pm as “the most popular time to tweet in any timezone.
The most commonly-used bot type both by pro-Ukraine and pro-Russia sides is the “self-declared bots — accounts that are transparent about being bots — “suggesting that authorities have identified these bots to be most useful in a information warfare campaign”. Self-declared bot accounts have the word ‘bot’ either in the username or bio.
The research also found the pro-Ukraine side to be using more astroturf bots than the pro-Russian side. Astroturf bots are hyper-active, political bots, continuously following other accounts to increase follower count of that account and/or systematically deleting content from their own account.
How bots trigger emotion
The research studied the words that most commonly appear in the bot accounts to note that “self-declared bots drive more angst about governing bodies. From a pro-Russian perspective, this may be to cause more disruption in the West, and from a pro-Ukrainian perspective, this may be to cause more disruption in Russia”.
The research paper observed that the bots also trigger angst by using angst related words, a majority of which are “surrounding fear and worry”.
The researchers therefore argued that bots and automated accounts “combine to increase fear in the overall discussion of the Russia/Ukraine war”.
The bots also increase online discussion around motion, observed the research paper, by tweeting posts with words like “moving”, “go”, “going”, “leave”, which are potentially associated with staying or fleeing the country.
Combining this with increases in angst suggests that bots could be influencing people’s decisions surrounding whether to flee their homes or not, claimed the paper.
(Edited by Poulomi Banerjee)