From a graphics/mod programming workshop, modifications of “Breakout” in awesome video form:
It uses Processing, a framework designed for animation and graphicky things. It was also used for the Similar Diversity visualization that maps out named entities and their common verbs across religious texts. It’s pretty cool:
… so a name gets bigger with its frequency, I get that; but what governs the size of an arc connecting two names? They say “common activities” — I’m guessing a distance measure on the terms in the contexts in which they appear? I’d be most interested in a graph of their explicit interactions; something like a count of co-occurrences in a proximity window (word radius? verse? chapter? book?) might do the trick…
This came out a while ago, I think; here are some related works posted on infosthetics.com.