As some of you know, I had a working relationship with Facebook while living in Mountain View, California to analyze click behavior of some of its users. Obviously, I am not involved with them anymore. There is alot happening behind-the-scenes that not many know about. I will not go into great depth over everything but just some of the general things people have asked I thought I'd share. As a disclaimer: Much of what I am writing is based on published scientific research I analyzed and public records. Facebook's spam prevention features appear quite simple but are actually based on machine learning and networking theories. Over the years, Facebook has determined to take advantage of stronger connections correlating to stronger networks. Therefore the goal over the next few years is to bring people into more networks with each other. For example, Facebook Places (launched) and the new Groups (launched) are allowing for more people to network with each other. There are several ways Facebook automation regulates spam control through user behavior. If you activate the rules of thumb below it can positively or negatively affect your account. You are likely to be labeled by Facebook as "potentially" bothering or sending unsolicited correspondence if you: send 4 friend requests and 3 are ignored send 40 messages in the span of one hour in a day send 10 wall posts in the span of one hour in a day not networked (friends of friends, school, work or location) You can be verified by Facebook if you: have an account from 2008 or earlier (only removes captcha) accept 20 or more friend requests (only removes captcha) submit a valid credit card on file submit a mobile phone address register from either a .edu or .gov e-mail address You are likely to have your account disabled by Facebook if you: get flagged by 30 or more users on a wall post get flagged by 10 or more users as "I do not know this person" on a friend request attempt to log in over 50 times in 10 minutes (incorrect password) directly impersonate or mispresent your identity (account or page creation) have more than one account If spam prevention is triggered, you will have a employee working in Facebook's User Operations manually investigate what is going on. They will determine whether or not you are being abusive at large (on a page, group or event) or privately. If your account becomes disabled, you will need to contact Facebook (User Operations) who will manually evaluate your situation. In addition, they work with software engineers, fraud investigators, and inside salesmen. Facebook keeps track of alot of data in order to analyze it for advertising purposes. One of the smaller departments in Facebook is the inside sales department. They talk to clients (advertisers) in order to sell space on the Facebook platform. If you notice, Facebook usually keeps track of your upstream (upclick) and downstream (downclick) behavior since its particularly useful for advertisers. You will notice this if you click on one page that is on another page. Since Facebook wants to retain their client to help bring in close to $1 billion in revenues, they work pretty hard to remove spam and keep order on an otherwise Wild West on the Internet. To finish up, there are plenty of interesting statistics about Facebook, its users and fan pages. Did you know that Thursday afternoons are the best time for user interaction (likes, wall posts, etc.)? Or that, fan pages catagorized (community excluded) as Website or Humor have 234% more users (historically) than pages on politics? Or pages on sex have 538% more interactions than pages on education? Finally, as I have been saying the last few months (before new groups, places came out, and the new spam filters) expect to see alot of changes and new things ahead. You can bet that these will be based on networks, connections whether friend-of-friend, similar likes (pages), or IP filtering on supercomputing.