Truthy Sleuths Track Twitter ‘Turfers

The Stephen Colbert-inspired Bloomington group at Indiana University has crunched massive amounts of data from the Twitterverse to track the spread of memes, and has found a number of accounts that aren’t playing fairly.


The Twitter sleuths behind “Truthy,” a meme-tracking website launched last month, are succeeding in their quest of rooting out so-called “astroturfers.” These are the types who stage PR blitzes disguised as genuine grassroots movements. It’s a pattern especially common during an election season, which explains the Indiana University Bloomington group’s timing.

By crunching massive streams of data from the Twitter site, the Truthy team has identified plenty of suspicious behavior. For instance, an account called @PeaceKaren_25 has fired off over 10,000 tweets since this summer, almost all of which support Republican candidates, particularly U.S. House GOP leader John Boehner. Another account, @HopeMarie_25, which was created 10 minutes after @PeaceKaren_25, retweets every single one of @PeaceKaren_25’s tweets without producing any original tweets. IU Bloomington School of Informatics and Computing Associate Professor Filippo Menczer thinks it more than likely that the two accounts are controlled by the same entity. (Whoever’s behind it, why stop at two accounts? @ChastityStacey_25 is probably available.)

The researchers say that these two accounts have become so prolific that they’ve created “Twitter bombs,” wherein Google searches for “gopleader” turn up the tweets in the first page of results. (When I tested this, however, those accounts didn’t turn up for me; Menczer tells me that the results vary depending on when the accounts are active.)

The Truthy team also uncovered a network of 10 bot accounts (that is, ones employing software to generate a flood of tweets) dedicated to smearing Democratic Senate candidate Chris Coons of Delaware. “Although many talk about Twitter bots,” Menczer tells Fast Company, “as far as we know this is the first time that they are being identified automatically and in real time.”

In a press release, Menczer’s colleague Bruno Gonçalves explains how the bots manage to avoid detection by Twitter: “duplicate tweets are cleverly disguised
by adding different hashtags or subtly tweaking the web addresses.” Though that gives the illusion of many different people sharing a single point of view, “in reality the bots are flooding the Twittersphere with one coherent political message.”

The Truthy team points out that many of these accounts appear to be in violation of Twitter’s terms of service. We’ve asked Twitter if they plan to take any action, and will update when we find out.


One of the coolest features of the Truthy team are the graphical representations of “diffusion networks,” which we wrote about earlier. Truthy has recently posted an interactive graphic with a number of the memes it tracks (the image at the top of this post is gleaned from that site). A healthy, natural discussion is represented by star-like explosions of blue and orange. A spam campaign is a giant block of blue, with just a few accounts retweeting each other. Interestingly, the #gop hashtag yields a diffusion network, pictured here, that looks like a twin star–representing the fact that there are two different communities, liberal and conservative, that use the tag in different ways, more or less without communicating with each other.

Update 10/27: Twitter spokesman Matt Graves says Twitter will review the accounts in question, and reminds us that any user can flag accounts as suspected spam from the spammer’s profile page.

About the author

David Zax is a contributing writer for Fast Company. His writing has appeared in many publications, including Smithsonian, Slate, Wired, and The Wall Street Journal.