by Glenn Caleval
If you are a friend or follow me on Facebook or LinkedIn, you will have seen the cautions from the Secure Surfing Organization about making use of particular “free” services. You may have personally sent me various invitations, virtual gifts or other service addons to share the social networking experience. And you will have noticed I almost always ignore such interactions, mindful of my own privacy and secure surfing habits.
Recognizing that a great many services are linking, inter-linking and exchanging user account information “to make things easier” for users, let’s look at how profiling works in simple language.
For this example we are going to consider
User1, Melissa, who has
- a Facebook account with 2 friends Susan, John
- a Google account with GMail
- a Technorati account
User2, Nicole, who has
a MySpace account
a Windows Live account
a LinkedIn account with 2 Connections, Frank, and Jill
First, just within Facebook
Before Melissa interacts with anyone else, she creates a profile database for Facebook. She includes her birth date, her favourite books, songs and movies; signs up to Follow the Misery Sings pop band and joins the Coping with Depression Cause. In the first few days she adds a bunch of game apps Solitaire, Slots Casino, Breakout Gems and so.
Before she has added any Friends, Melissa already has a deep psychographic profile built that thinks it knows the following kinds of things about her:
*This is for explanation and may not reflect the physical technology used to draw the conclusions*
A formula assigns different values to different profile data item. A single profile item may have more than one value. For example because Melissa entered in her favourite books, a value is set that shows she reads books; because she listed four books, her book-reading value is set to high.
One of the books she included is Titled “Momma and the Meaning of Life” and ISBN database catalogues that title under the subjects of “mental health” and “psychology.” (ISBN is the International Standard Book Number assigned to every published book and catalogued by subject, author and so on). This data item may have a value related to mental health.
Her favourite song is “He stopped loving her today” by the immortal George Jones. She gets one value for Country Music, another for “sad music” and combined with her favourite book a new value is set, the “loneliness value” initiates at a modest strength of 2.
She is following the band Misery Sings. She gets a new value for Pop Music and her loneliness value increases to 4.
She joins Coping with Depression and her loneliness value skyrockets to 9. With the games she signed up for all being single-player, the loneliness value jumps again to 13 and when after four days she has added no friends it is increased to 15.
So what the heck good would a “loneliness value” be to Facebook; why would they calculate such a thing?
First, be clear that I am not claiming that Facebook or anyone else actually has a “loneliness value.” They have other “algorithms” or formulas to calculate the various deep characteristics that they can exploit with advertising or up-selling.
A high loneliness value would be useful to target self-help books, lure into artificial environments that can substitute for human company, target political messages, or impel donations. Knowing a person is vulnerable in specific ways is the most powerful tool available to manipulate behaviour including buying decisions and social and political choices.
In this case, they would all be off-point because Melissa, you see, is a second year psychology student, she doesn’t suffer from depression but likes to support those who do, “Mama and the Meaning of Life” was required reading in her class and she just like the book, she thinks “He stopped loving her today” is one of the highest camp set of lyrics ever written and her boyfriend plays in the band Misery Sings.
She hasn’t signed up any friends because she wants to get some experience with Facebook before exposing herself and the games she chose allow her to do that without worrying about mistakes in etiquette or social rules.
So across the board in the case of Melissa, the formula is just wrong. All the same, she will exist in a major database as a depressive and that value will be exchanged in many ways that will create a “systemic knowledge” of Melissa as a depressed, lonely 22 year old female.
Now think through all the other kinds of calculations that can be done with all the other choices, applications and information you share on Facebook.
After a week or two, Melissa is ready to invite her first friend, Susan.
Susan is Melissa’s sister so Melissa uses the MyRelatives app which Susan accepts so Facebook records them as sisters.
The profiling formula operates with an assumption that relatives are likely to be more alike than different, so all the psychographic values created in Susan’s account are given higher weight in Melissa’s account that if they were not relatives.
Susan is immediately told by Facebook to suggest other Friends for Melissa and feeling some pressure to be a good sister, Susan sends her own boyfriend John a suggestion to befriend her sister, which he accepts. Now John’s values are added into the calculation. Because Susan and John complement one another with many similar traits, the addition of his data values boost those scores for all three. At the same time, there are a few items in John’s profile that give him a loneliness score too and it is inflated through the friendship suggested by his girlfriend.
Part 2 coming soon.
