George Mikros Profile picture
Professor of Computational and Quantitative Linguistics at Hamad Bin Khalifa University at Qatar. Research on Computational Stylistics

Mar 6, 2022, 15 tweets

1/12 This thread is about pro-Russian propaganda in the Greek Twittersphere inspired by my colleague @marcowenjones research on disinformation war currently unleashed on Twitter. The hashtags analyzed are "Με_τον_Πούτιν" "Με_την_Ρωσία" #πολεμος_στην_Ουκρανια #UkraineRussianWar

2/12 From 19 Feb 2022 till 5 Mar 2022, there were 596 tweets posted by 85 accounts using these hashtags. 51 tweets were isolates meaning they were not engaged in any interaction. The average rate of tweets/day was 9 since the beginning of the war, with low daily variations.

3/12 What are these tweets talking about? Here is a word cloud of the 100 most frequent words in this corpus. There are positive words related to Putin as leader, negative words associated with the Greek government's engagement in the conflict, and NATO's and USA's negative role.

4/12 I used a more sophisticated quantitative vocabulary analysis (t-SNE) to explore the hidden semantic dimensions under these tweets. The analysis revealed 3 major clusters of topics in the pro-Russia tweets that can be seen in the following image.

5/12 The first words cluster (green circle) is about accusations to the Greek government. The second cluster (blue circle) mainly focuses on the US and NATO role in the conflict. The third cluster (pink circle) contains words of admiration to Putin.

6/12 To dig further into the collected corpus, I performed a network analysis excluding the isolates. I plotted the accounts interactions, and using a clustering algorithm (Clauset-Newman-Moore), I identified 4 main groups of accounts that post using the pro-Russian hashtags.

7/12 I identified the influencers in the plotted network structure using the Betweenness Centrality, a measure that quantifies the potential of a node to control the information flow in the network. The more important is a node in the information flow, the node turns to red.

8/12 In the network diagram, we can see that the most influential account is the Russian Embassy of Greece (rfembassygr). Nearly all the nodes refer to this account, as you can see in the highlighted connections diagram below.

9/12 The account of the Russian embassy is also connecting all four clusters of accounts since it seems that all groups are mentioning or retweeting their tweets. Each cluster has different central influential accounts, but all of them share the same vocabulary discussed above.

10/12 I didn't detect any pattern of automated bot activity (e.g. synchronized posting or similar content posting in a short time). However, I found central nodes in each cluster that exhibit atypical behaviors, e.g. the account deadwarrior_ that suddenly disappeared yesterday.

11/12 Another interesting central node is proedrostrump24. It joined Twitter on 11 Feb 2022, and in 23 days, it posted 1035 tweets most of them related to the Ukraine – Russia war.

12/12 This is just a small contribution to understanding social media's role in modern warfare. Disinformation techniques are also being used in the Greek Twittersphere, and our only defense is the critical and multiview approach of the incoming news.

11/12 Another interesting central node is proedrostrump24. It joined Twitter on 11 Feb 2022, and in 23 days, it posted 1035 tweets most of them related to the Ukraine – Russia war.

12/12 This is just a small contribution to understanding social media's role in modern warfare. Disinformation techniques are also being used in the Greek Twittersphere, and our only defense is the critical and multiview approach of the incoming news.

Share this Scrolly Tale with your friends.

A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.

Keep scrolling