Below is a graph of my Facebook network, created by the very nice Friend Wheel application which draws a customizable, brightly colored network graph of up to 400 of your BFFs. I had a grand time annotating it on Flickr.
You can ask it to graph a subset of your friends, say one of your networks, or all your friends in common with another person. My clusters are fairly discrete, though not as much so as, say, this person’s, and I’m not sure how much of the difference is due to constraints on well-grouped the names can get.
Are there meaningfully different visual patterns associated with different personalities and social settings? I experimented with the latter by graphing the following subsets of my Friends List:
1. Graduate school friends. The Mafia in fine form. Just look at that rainbow of community love.
2. Grade school friends. Mostly clustered by grade, just as it was back in the day.
3. Friends in common with my brother, who’s the blue-red node at lower right on my full Friend Wheel. Not only do we have lots of friends in common, but they’re friends from several different groups – making my brother a dense node unto himself rather than part of one of the larger clusters. You can see these different groups in our mutual-friends Friend Wheel below: the clusters correspond to various sets of school friends, work friends, and Bangkok friends. The lower half is our high school.
See Flickr for other examples of Friend Wheels.