There’s a great piece on Nieman Journalism Lab about Jacob Harris’ hatred of word clouds. Harris, a journalist and software architect for The New York Times, argues that in terms of data visualization, word clouds just don’t deliver any valuable information. He gives some great examples of articles that use word clouds versus articles that use other more successful forms of data visualization.
I agree with Harris about word clouds popping up too frequently in journalism, but think that word clouds can be used as an online self-promotion tool for freelancers. If you go into your resume and remove all the non-descriptive words and leave all the action verbs and descriptive nouns you can turn that into a word cloud resume. I wrote a post about self-promotions for freelancers that describes this idea further.
Perhaps Harris would condone my use of a word cloud resume. He concedes that word clouds are useful for textual analysis. For example, if a Phd student wants to show how many times an author uses a specific word in a great work of literature (although arguably you could use a simple chart.)
Harris says that the biggest problems he sees with word clouds, is when a news organization uses a word cloud in a situation where textual analysis is not appropriate, like in a war analysis. Seeing how many times Iraq War coverage mentions the words “car” or “blast” really doesn’t give readers any new insight into the conflict.
Here is one choice passage from Harris’ post where he compares reading word clouds with looking at tea leaves:
I’ve seen this pattern across many news organizations: reporters sidestepping their limited knowledge of the subject material by peering for patterns in a word cloud — like reading tea leaves at the bottom of a cup. What you’re left with is a shoddy visualization that fails all the principles I hold dear.
The best part about Harris’ blog post is at the end, when he taunts the “sadistically inclined” readers who may go ahead and make a word cloud out of his post with this.