Tag clouds are a nice way to visualize the content tags of a website. Flickr started this trend when they displayed a “All-time most popular tags” tag cloud on their front page. The size of the tags in the tag cloud is usually relative (more frequently used tags are displayed in a larger font). In this way you can quickly see what is hot and what is not on a webpage.
But when you add the dimension of time things get really interesting. You can actually map out shifts in tag usage. Chirag Meta created a Tagline Generator “that lets you generate chronological tag clouds from simple text data sources without manually tagging the data entries.” Two nice applications of this generator are:
- US Presidential Speeches Tag Cloud
You can see a shift from “welfare” to “terrorism”
- Microsoft’s evolution, in keywords
The word “computer” is disappearing, interesting!
It might be nice if you could focus on one word, or a set of related words, so you could follow a certain trend. What else can we do with tagclouds (except making a t-shirt out of it of course)?
3 thoughts on “Tag clouds as a research object”
The one about Microsoft is pretty well done, for the upcoming MA thesis I’m using a kind of tag cloud to have clear what is the most important. Just to have an overview. So that’s kind of the other way round.
Isn’t that the time way? Or am I misinterpreting you?
Ehm, yea I guess you could call it the time way :)