I constantly find high-volume accounts in certain hashtags and they are present in #QAnon networks. These high-volume accounts are hubs in Twitter networks. Here’s what some of the high-volume hubs look like in 8 days of #QAnon tweets from May 3 to May 11 medium.com/@erin_gallaghe…
In network science, hubs act like bridges between smaller nodes & help spread information faster. In the case of a network of people with a contagious disease, a person who has contact with many people is a hub which spreads the infection widely. medium.com/@erin_gallaghe…
Hubs in a network make the network stronger. Random smaller nodes in a network can fail (or in this case their accounts can be suspended) and the network will remain intact, still connected by the larger hubs.
User-to-user graph of 48,241 #TheStorm tweets - April 27 to May 17
I started a rolling twitter thread of what I called #MAGA spammers, many of these accounts are also tweeting #QAnon hashtags (along with other MAGA related trends)
User-to-user network of 18,868 tweets mentioning @PradRachael from May 16 to May 21
The @AP recently reported on 2 accounts that get mistaken for #bots but are actually people who are members of Twitter rooms. @MAGANinaJo is a 70 year old grandmother who is tweeting 73 tweets per day. medium.com/@erin_gallaghe…
16,232 tweets mentioning MAGANinaJo — Aug 5 to Aug 8
Here's the description of Twitter rooms in the @AP article apnews.com/06efed5ede4d46… Communities of people coordinating their tweets.
If Nina is participating in Twitter rooms with other high-volume accounts and they are retweeting each other’s tweets, then their public tweets will be connected even though the rooms are private.
High-volume hubs in 16,232 tweets mentioning @MAGANinaJo — Aug 5 to Aug 8
have to charge my laptop awhile brb
I found a conservative political marketing firm called @AppSame in @MAGANinaJo's network. It's the top account in the influencer index for her account, meaning it's the account w/ highest number of followers mentioning her. AppSame has 370,819 followers & tweets 107 times per day
Looking back through the datasets I previously collected, I found @AppSame in the top of the influencer indexes for #QAnon, #TheStorm and @PradRachael's account. A conservative political marketing firm is connected to all of these networks.
AppSame's website doesn't say much about them. Their recent posts are for essay writing services. I found a phone number & address on LinkedIn. The address is a shopping center in Tampa, FL & the phone is a Google voice number. I called, no answer. medium.com/@erin_gallaghe…
AppSame has one employee publicly listed on LinkedIn: Eric Stole. I found an “Eric Stole (AppSame)” on Facebook. Per LinkedIn, Eric went to U. of Tampa & previously worked @ Oracle but on Facebook he went to Holy Angel University in the Philippines & previously worked at Facebook
Lots of his Facebook friends also seem to be located in the Philippines. 🤷 No idea if Eric is real or not but of course I was curious what the network surrounding @Appsame looked like.
Here's a user-to-hashtag network for 2,477 tweets in @AppSame's network from Aug 8 to Aug 9.
Here's the user-to-user network of 2,477 tweets mentioning @AppSame from August 8 to August 9, about 12 hours worth of activity. There are high-volume hubs in the large orange cluster. medium.com/@erin_gallaghe…
Those activity patterns were caused by tweets mentioning @CinepolisUSA, a theater in FL that temporarily suspended @DineshDSouza's new movie last night. Several high-volume accounts are tagged in the tweet and it received thousands of likes & retweets. medium.com/@erin_gallaghe…
I don’t know if this was organized brigading of @CinepolisUSA from a Twitter room but networks of high-volume accounts and communities of people coordinating in Twitter rooms *could* be mobilized to target certain accounts or hashtags. medium.com/@erin_gallaghe…
Several reoccuring components in these networks: high-volume hubs, a political marketing firm and Twitter rooms with real people. It would be beneficial to include real people because they would have a real phone number to add to their accounts & overall seem less bot-like.
Twitter posted a blog on June 26 that seems to address these kinds of high-volume accounts. I've been waiting to see if any of the accounts I've been tracking would be suspended but so far... nope. blog.twitter.com/official/en_us…
So far I tweeted 63 high-volume accounts with at least 10K followers each who tweet 100+ times per day. Since some of these accounts are in @MAGANinaJo's network & @MAGANinaJo is in Twitter rooms, it's likely some of these accounts are in Twitter rooms too
Ok I think that's probably enough tweets for tonight. Apologies for the monster thread, I tried to be succinct but it's a lot to unpack. Hope it helps!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
OK my @medium account got "caught in the spam filter" & has been restored so let me try this #thread again. I've been working for several months on a series of blogs explaining trolling & botting tactics used by 4chan during the election. #ContentWarning: medium.com/@erin_gallaghe…
Several months ago @elimisteve from @PursuanceProj found a series of pastebins and sent them to me. 4chan trolls studied learned graphic design software to make memes, memetic warfare and botting to troll the election. medium.com/@erin_gallaghe…
The 1st part is an 8chan thread by someone who claimed to have worked in the marketing & advertising industry. The explain various components of effective graphic design & instructions for /pol/ to “normalize” their message to blend in with the normies. medium.com/@erin_gallaghe…
1/ #WalkAway#FakeNews#Thread Michele Stapleton posted these posters (and the stock images they came from) on Facebook and seeing them all together I recognized the style guide.
2/ couple more... I have seen this font on propaganda posters before.
3/ during the 2014 Gaza War we had so many trolls on the @NewsRevo FB page I made a blog about the 2009 Hasbara Handbook (now offline) and included screenshots of the posters that were going around then. Here's the blog in the wayback: web.archive.org/web/2014101705…
2015 UNHCR study "Women on the Run" published first hand accounts of refugees fleeing El Salvador, Guatemala, Honduras and Mexico acnur.org/fileadmin/Docu… (60 page pdf)
"Women interviewed for this report indicated that they and their children face extreme levels of violence on a near-daily basis. They described being raped, assaulted, extorted, and threatened by members of criminal armed groups, including gangs and drug cartels."
"Sixty-two per cent of women reported that they were confronted with dead bodies in their neighborhoods and a number of women mentioned that they and their children saw dead bodies weekly."
Just poking around this interesting website... "TweetAttacksPro 4 is designed to run thousands of twitter accounts at the same time 24/7"
Unlimited accounts that you can automate 99% for $216 per year. 🙃
It's a full service website. Manage and automate an unlimited number of Twitter accounts and run a campaign right in the dashboard.
Here is the problem with attributing bots to a certain location or person. Let's say for example you buy 1000 Twitter accounts to run a campaign and you bind them all to a proxy in Canada. Are those Canadian bots?