Discover and read the best of Twitter Threads about #ميدان_التحرير_الان
Most recents (3)
Trying various color palettes to highlight the spam... here's the user-to-hashtag network for 17,049 #ميدان_التحرير_الان tweets from Sept 21 until this afternoon. White background w/ labels & spammers highlighted in cyan and on gray w/o labels & green. They look like tumors. 
will update my blog with these images in awhile: medium.com/@erin_gallaghe…
I isolated the spammers and made a user-to-hashtag graph of the trends they targeted. This shows 2741 tweets sent by 38 accounts. 37 of the accounts used @IFTTT and one used Twitter Web App (a now suspended account called ____6HH0). Spam was ~16% of my dataset of 17,049 tweets.
Hashtag visualizations of #ميدان_التحرير_الان - Fake Twitter accounts bombarded trending hashtags with automated spam while protests were breaking out on the ground in Egypt in what appears to be a deliberate campaign to disrupt the protests. medium.com/@erin_gallaghe…
The spam campaign is still active this morning. This tactic effectively disrupts the flow of real-time information online. Look at "latest tweets" for the trending topics right now in Egypt. Instead of relevant tweets, you'll find these repeaters flooding the hashtags with spam.
Found 4 more just now in the "latest tweets" for hashtag #انزل_انت_مش_لواحدك which is currently in the top trends for Egypt. All using @IFTTT to post spam.
[#Thread] 1/ This one is on the Egypt protests, or rather the hashtag #ميدان_التحرير I gathered a sample of around 21,000 interactions from around 9,700 unique accounts. Quite a lot of accounts considering they are unique. Obviously there will be much chatter about influence ops
2/ I did my usual analysis first. I examined the account metadata to determine the distribution of account creation dates. What I found was quite astonishing. 1484 of the 9713 unique accounts were created in September 2019. That's 15%!
3/ As you can see from the graph, the spike on the right shows the 1484 accounts created in September. Bear in mind the average number of accounts created per month in this sample is just 66. That's a huge difference! What's perhaps more interesting is that the accounts started
