This morning I was inspired to do a bit of word cloud exploration ...
I was prompted by a general tweet from Kevin Lim about "making beautiful word clouds using wordle.net. He had constructed a word cloud based on his del.icio.us entries ...
Kevin, in turn, was prompted by Henry Farrell's posting of his text cloud based on his book “The Political Economy of Trust: Institutions, Interests and Inter-Firm Cooperation” ...
Henry was reacting to Steven Poole's posting of the word cloud based on "Unspeak" ...
So, what does it all mean?
At the simplest level, a "word cloud" ("text cloud", "tag cloud" in the case of a blog) displays words whose font size is proportional to that word's frequency of occurrence in the analyzed text. It's a graphical concordance.
The placement of the graphemes, however, is not so rigid. Some clouds display the words according to size (boring :)), some alphabetically (not as boring), and some use algorithms that are not always published (interesting, since one is left to guess the algorithm).
The process can be made more analytical and comparative ... the IBM's Many Eyes project can compare two text clouds (recent examples were campaign speaches by Hilary Clinton and Barack Obama ... expect more of this genre this fall).
Henry Farrell, Wordle, http://crookedtimber.org/2008/07/21/wordle/
Kevin Lim, Wordle: Make beautiful textclouds…, http://theory.isthereason.com/?p=2285
Many Eyes, http://services.alphaworks.ibm.com/manyeyes/home
Steven Pool, Can Freedom Possible, http://unspeak.net/can-freedom-possible/
Wordle, http://wordle.net/
No comments:
Post a Comment