Over the weekend, I read an interesting article, “Edmund Burke: How did a long-dead Irishman become the hottest thinker of 2010?“, by Amol Rajan in the Independent on the philosopher, Edmund Burke. In the past I’ve read his musings on the sublime in “A Philosophical Enquiry into the Origin of our Ideas of the Sublime and the Beautiful” and seen his “Reflections on the Revolution in France” on the shelf but have yet not pushed myself to exploring it. The article that I read changed that as it managed to draw out the relevancies of his argument with the current ConDem / LibCon coalition. It was written as a response to David Marquand’s article, “Patron saint of the big society” in Prospect magazine.
I still have not read the Reflections yet but I thought it might be fun to get the Gutenberg text, strip it down to the text itself and begin building a tag cloud as an exercise in text mining. What I’m trying to achieve is a visualisation of the text and the keywords. That will probably take me a couple of attempts in cleaning the information up and making it relevant.
Having done that, I’d like to make the cloud relevant by clustering words together and linking them into the text or search. That is one of the things that really annoys me with some tag clouds. Pretty pictures but no linking. I always think of it as why bother? Seriously, time is taken to mine the data, then visualise it but not to make it relevant or link it into the source material somehow.
For now, I’m going to leave it as a pretty picture version (sorry…) as I try to make it more relevant and release bits of it as I get it done.
Tag cloud of Burke’s Reflections on the Revolution in France.
On the todo list is linking the words to either a search and text and to contextualise by creating a corpus of words.
This follows on some stuff that I was doing a while ago to explore how make tag clouds and then use them. I was reading Jim Bumgardner’s Building Tag Clouds in Perl and PHP (OReilly, 2006). Lots to do but something to chip away at.
Update: Updated the Prospect link which is now available and corrected a spelling error