Clouds of words
15 January 2006
Another nice visualization tool by Jean Véronis, published on his extremely interesting blog about language and technology. This one does data mining on results of a french-speaking web search and presents the words most often associated with the one you entered, to produce a cloud in which size and color of the results represent the statistics of association. It's called "Le Nébuloscope". I only changed the colors to something that fits me better. I think I already saw something like that - including similar tools for bio-informatics - but it's nevertheless very nice to be able to play around with this kind of tool. The power of certain types of visual presentation is amazing. Let's check what I am associated with on the french-speaking internet. First with my nickname, and then my real name :
Well. It looks like I am a researcher and I had some trouble with the law because of something related to computers and anti-virus. Damn. That's quite true. And I thought I was famous for my t-shirt collection.
You can use the same visualisation engine with a different original database. For example, a one-hour long talk by some politician can be represented with this same tool, like here. You can see in two seconds what were the main points of the talk. On the other hand, it's a bit frightening. Maybe in the future students will learn what "War and Peace" or "Les Misérables" or "Alice in Wonderland" are about, only by looking at one of these clouds of words for two seconds.
[Note to myself : it's quite stupid to write in english about a tool for searching french-speaking texts]