So, according to the latest viral meme, I belong to two cliques of size 9:
I am a member of 2 cliques of size 9
- tacit, anklesnake, lightgatherer, smoocherie, moonshadowdance, datan0de, femetal, zensidhe, fatesgirl
- tacit, lightgatherer, kellyv, smoocherie, moonshadowdance, datan0de, femetal, zensidhe, fatesgirl
This is interesting for its own reasons (there are some common denominators in all the names listed, which I’ll leave as an exercise to the reader to describe), but I think what’s more interesting is how it works, which is described here.
Discovering these two threads of nine users in my friends list required a couple of minutes of activity on a server of unknown capacity and about 6,000 database search steps, and turned up lists of people with several significant things in common with me. Yes, it’s a toy, but it also demonstrates the power of data mining.
Now, let’s suppose the same thing were to be done to, say, social or political networks in government databases. Let us further assume that the government were looking for people with, say, subversive political views. I’m sure you can figure out where this is going…
The software is dumb. It knows nothing about me and it knows nothing about the people on my friends list, yet it teased out a list of people with very unconventional social views, based on nothing more than the fact that like-minded people tend to gravitate toward one another. The software doesn’t have to be smart in order to be able to reveal patterns of social activity that a human observer might not see.
So. Fred is in a government database. Fred is being scrutinized by the government, because he has Dangerous and Unacceptable Ideas–that Iraqshould be left alone, say, or that George W. Bush is a miserable failure. Or perhaps he’s a Communist–who knows?
So. The database plots Fred’s social contacts–who he talks to, who he goes to the movies with, who he has lunch with. The government wants to find a list of other subversives like Fred, but it doesn’t want or can’t afford to spend the manpower to knock on all of Fred’s neighbors’ doors. Well, hey! Who needs that, when all you really need to do is search the database of social contacts?
As a LiveJournal meme, the cliques list is kind of neat. As a proof-of-concept for social database mining, it’s slightly disconcerting.