Facebook Explained

tronlegacy

Junior Member
Joined
May 21, 2010
Messages
121
Reaction score
47
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.
 
Can you tell more about the role of IP addresses in spam protection?
 
This certainly explains a few things I was wondering about. Lol
 
Nice info ,,, I was reading something similar yesterday but about twitter ..
 
HTML:
attempt to log in over 50 times in 10 minutes (incorrect password)

That is wrong. With 4 incorrect attempts to login to a facebook account. The account then becomes locked.
 
Can you tell us anything more about the use of private proxies and what kind of role they play?
 
Better yet, what's causing the spam filters to kick in while posting on walls or sending messages through IM feature.

Is it using free redirects, something in the phrase of your message? I wanna know what triggers facebook blocking your link or message.
 
Thanks Tronlegacy! Can anyone confirm these behaviors by FB by the way?
 
just tried the 3/4 accepted request one and now I'm getting "do u kno this person"
yea i had my account shut down after sending too many private messages
i also registered as a famous figure's accounts and it got shut down after 3 days
 
Thanks for sharing your insights tronlegacy, but your list under "You are likely to have your account disabled by Facebook if you" is no longer the case, far from it. I'm sure it was when you worked there, but not anymore.

"If spam prevention is triggered, you will have a employee working in Facebook's User Operations manually investigate what is going on."

500m accounts, 1600 employees, 1000 of whom are in security. They don't read messages. It's impossible. They ALWAYS go with the complainer. You hit 'spam' on someone and their gone.

"If your account becomes disabled, you will need to contact Facebook (User Operations) who will manually evaluate your situation"

I have never heard of anyone getting in contact with a human being over there let alone getting their account back. But that's just me.
 
I have never heard of anyone getting in contact with a human being over there let alone getting their account back. But that's just me.


They did contact me personally, and they revived my account. But that's only after I promised them I would buy thousands of dollars in FB ads.
 
They ALWAYS go with the complainer. You hit 'spam' on someone and their gone.

Not true at all. I even tested in one of my own accounts by reporting it 12 times with different accounts and nothing happened. Facebook has actually become more lenient. I remember spamming the facebook development blog (of all places) and nothing happened.
 
What about inviting people to a page? Even if its legit business pages, can you be triggered as spammy?
 
Back
Top