Wordle and Tag Clouds in the Classroom

Posted by Danny on Thursday, June 18, 2009 at 11:13 pm

Dad was a devotee of computers probably minutes after discovering them as an undergraduate at Waterloo in the 70s.  He repeatedly tried to instil the same wonder and excitement in me, groping for ways to connect the nature of computing machines to my own interests and probably disappointingly artsy foci—anything, at least, to extend their significance beyond the video games I was playing.  One of the early uses, I learned, was for academics studying literature to compute word frequencies in texts, which was for me at once a completely novel idea, and seemed spectacularly boring and pointless.[1]

I hadn’t thought much about it until “tag clouds” started popping up on popular sites and the possibilities of this kind of data visualisation started revealing themselves despite my benightment.  Recently, Geoff brought Wordle to my attention. It’s a groovy online tool by Jonathan Feinberg for building CC-licensed, aesthetically pleasing weighted lists in the style of tag clouds using text from various sources: you can paste in your own text,[2] the URL of an RSS-feeding website (Philomathy.org and probably any modern blog), or a del.icio.us user account.

A bunch of applications for educators comes to mind.  If teachers request emailed copies of their students’ assignments (a practice I hope to soon become obsolete the way requesting papers to be typewritten is now), it would be a snap to make colourful collages of individual students’ or whole classes’ clouds.  This might be appealing to students in other classes (and their teachers) working on the same projects as a way to develop collegiality and interest in peer work and research before university.

Less idealistically, if teachers have access to the digital version of their assigned readings,[3] providing clouds to students as part of their introductory material may supplement or even replace summaries as effective priming tools.  I always read more efficiently when instances of recognition bolster my confidence—maybe teachers could use this to facilitate study or accommodate students with certain mild reading exceptionalities.  I’d love some day to hear from literacy specialist on the topic.

Also, creative, technical, and academic writers (post-secondary or otherwise) looking to hone their written composition skills could use Wordle to identify over-used words which they could subsequently avoid.  Users have the ability to include filters for “common words” by language which could be manipulated (crudely right now, or developed for finer control by Feinberg in the future) to reveal patterns pertinent to this sort of self-analysis.

Also, it’s fun!

(A tip for those looking to print their Wordle-clouds on large-format paper or manipulate them artistically: there is a print option built into the applet, but as far as I can tell, no way to save your cloud as an image file.  If you want to work with a high-resolution of your cloud, print it to pdf—the vectors are maintained—and then open it in Photoshop or Illustrator at an arbitrary resolution.  More fun tips and ideas are available at Feinberg’s FAQ.)

(Also also, the Wordle homepage displays clouds saved by random web denizens, usually intended to be artsy.  As Feinberg mentions, these aren’t moderated and can contain all sorts of words that could get teachers in trouble for directing their students to; if you’re thinking of using Wordle as a teaching tool and telling your students where to find it, be aware of your liabilities.)

---

[1] Though I’ve grown more speculative and systematic in my willingness to entertain ideas that seem initially batty, I suspect my inability to muster more than a vague, academic interest in the dusty concordances in Nipissing’s reference library was predictive of my further studies totally exotic the discipline of my B.A.

[2] I’ve tested it with the text from my undergraduate research project, so the field can accommodate at least 9500 words.

[3] This is a scenario I hope will become more and more probable, especially as OER and their distributive infrastructure worms into the mainstream.  I’ll write more about this soon.