1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Twitter takedowns revealed.

Discussion in 'Black Hat SEO' started by JustUs, Aug 16, 2013.

  1. JustUs

    JustUs Power Member

    Joined:
    May 6, 2012
    Messages:
    609
    Likes Received:
    452
    AN agency under contract with twitter conducted a study of Twitter spam and purchased Twitter accounts. The information revealed by this study was used to identify fraudulent twitter accounts, where they came from, the methods used to create them and some of the sellers of those accounts. The paper reveals that many of the fraudulent accounts were purchased here at Blackhat World. The paper names names and services that were selling those accounts, and sometimes shows how the same sellers, resold the same purchased accounts several times.

    The study is 16 pages, the title is "Trafficking Fraudulent Accounts: The Role of the Underground Market in Twitter Spam and Abuse, " and the link to that study is http://krebsonsecurity.com/wp-content/uploads/2013/08/USENIXtwitteraccountpaper.pdf.

    For anybody considering purchasing Twitter accounts, this paper is a "Must read." For anybody that creates those account, this paper is a "Should read."

    For those sellers that sell the same accounts: Shame on you.

    Happy reading.
     
    • Thanks Thanks x 5
  2. theadultsplayground

    theadultsplayground BANNED BANNED

    Joined:
    Jan 24, 2008
    Messages:
    1,202
    Likes Received:
    336
    2.2 Impact of Fraudulent Accounts:

    Todays word of the day is "Miscreants", lol.
     
  3. kvmcable

    kvmcable Supreme Member

    Joined:
    Dec 28, 2010
    Messages:
    1,355
    Likes Received:
    2,815
    Occupation:
    24 year business owner - old school dude
    Location:
    KFC - BW3
    Well that would explain the recent panic by some twitter sellers and resellers. Seems Twitter has hired some pretty smart people to create a very good algorithm picking off fake accounts. 99.99% is a success by most standards.

    I'd hate to be a Twitter account seller or reseller atm.
     
    • Thanks Thanks x 1
  4. dinkish

    dinkish Power Member

    Joined:
    Apr 19, 2013
    Messages:
    689
    Likes Received:
    159
    Hah, just bought a service off here 2 days ago.

    It was crappy, I'm glad I'm seeing those "permanent active followers" dropping off already. Didn't even have bio pics.
     
  5. CosmicSoundz

    CosmicSoundz BANNED BANNED

    Joined:
    Apr 30, 2012
    Messages:
    1,230
    Likes Received:
    1,296
    [​IMG]
     
  6. dheer

    dheer Jr. VIP Jr. VIP Premium Member

    Joined:
    Jul 24, 2009
    Messages:
    2,443
    Likes Received:
    1,029
    Home Page:
  7. Whisker

    Whisker Moderator Staff Member Moderator Premium Member

    Joined:
    Dec 26, 2007
    Messages:
    994
    Likes Received:
    1,322
    Narcs. Narcs everywhere.
     
    • Thanks Thanks x 1
  8. JustUs

    JustUs Power Member

    Joined:
    May 6, 2012
    Messages:
    609
    Likes Received:
    452
    If you read the paper closely, there are a few things that should stick out besides the math that a person could base an anti algorithm on. When the paper informs of how they detect those accounts, it also informs how to evade detection. Because this thread is public and the paper clearly identifies this forum as being on that the investigators hang on, I will not give what I picked out. The paper does, however, tell one, between the lines, how to evade the new algorithm.
     
    Last edited: Aug 16, 2013
  9. imserious

    imserious Senior Member

    Joined:
    Mar 27, 2009
    Messages:
    946
    Likes Received:
    560
    One thing to learn from this study:

    Break the pattern as often as you can.
     
  10. Asif WILSON Khan

    Asif WILSON Khan Executive VIP Premium Member

    Joined:
    Nov 10, 2012
    Messages:
    10,139
    Likes Received:
    28,607
    Gender:
    Male
    Occupation:
    Fun Lovin' Criminal
    Location:
    London
    Home Page:
  11. ShabbySquire

    ShabbySquire Power Member

    Joined:
    Nov 30, 2011
    Messages:
    574
    Likes Received:
    122
    Location:
    UK
    Reminds me of the white paper on FaceBook, and the methods they used to detect rogue activity.

    Like most algos that set out to thwart the undesirables, it effectively boils down to this: act like a human.