Posted by: rolfsky | June 20, 2010

10 tips for identifying fake twitter accounts

these tweets are from bot accounts, referencing events from the past as if they haven't happened yetToday, my wife replied to a tweet of mine and a few minutes later, I noticed an odd retweet of her response.

“@rolfsky The OpenOffice version is worse, though”

Without the context that I was wrestling with Microsoft Office’s bulleted list implementation, it didn’t make a lot of sense to RT. Something only a bot would do.

Another curious element was that the retweet by “unixland” was made with the Perl Net::Twitter package, not by any recognized twitter client. Suspicious.

A quick check of unixland’s feed shows a fair number of retweets that seem to have little common thread, and then liberal inclusions of links to “Computerhulp” in Amsterdam, and the owner’s personal photography website. Bogus.

This got me digging a little deeper, and I clicked on the hashtag for eBay’s recent developers conference, #eBayDC10. Bizarrely, though the conference had ended, there were several tweets about it not having happened yet. Bizarre.

10 Tips for Identifying Fake Twitter Accounts

  1. tweets seem to follow no central theme or narrative, perhaps referencing conflicting locales or geographies
  2. tweets reference items in the past as if they haven’t happened yet
  3. many tweets begin with an ascending series of letters (an attempt to fake uniqueness)
  4. first name is filled out as “Name” or username is in the form of “firstName_lastName” or a string followed by digits “fooBar1234”
  5. posting application is some type of automated or command line driven method such as just “API” or “Perl Net::Twitter” (needed for automation)
  6. many/most of the tweets contain links
  7. the recurring links lead the same place (or even the same tweets, repeated!)
  8. no web, bio, or location information
  9. are only following popular individuals
  10. are only followed by other accounts that have very few/no tweets

Of course, this is now a cat-and-mouse arms race, escalating the conflict between the virtuous Twitterers, and the despicable spammers that want to profit off everything. And even writing down this list will help the spammers, as a number of these items are easily resolved, such as adding in additional stolen information for web, bio, and location links, creating more varied tweet schemes, and automating via the web instead of API.

I say, bring it on spammers. (Because at the very least, that will create at least one anti-spam job position at Twitter.)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: