Twitter Mining

A few weeks ago I was talking to Phil about the rich data that could be found in blog comments. We thought that aggregating the list of frequent posters over time could actually give you an interesting picture of who reads and comments on which blogs. Some interesting conclusions could be drawn, and some interesting friend graphs could be generated.

Looking at the open and freely available Twitter friends list for John Edwards I see yet another interesting place to collect some interesting data. You could hazard a guess that these people are Democrats, politically active, and so forth. Even better, intersect a couple of lists, do some pivoting, and who knows what you might learn.

Clearly this isn’t unique to Twitter, but the fact that Twitter makes it so easy to express your interest in following someone, combined with the public nature of the data, certainly makes the raw data a lot more accessible.

By the way, my Twitter page shows names because I am running Wes Maldonado’s Greasemonkey script.