Oct 26

How do we get content from word clouds?

Some time ago I became involved with a primitive game called the Geowiki game that was being studied by Clair Dormann a postdoctoral fellow in the HOT Lab at Carleton. The work was being done in conjuction with the Cybercartography initiative.


The game involved sequentially picking three commands, one from each of three categories. The participants would then record their observations made while carrying out the commands. For example, groups of students from a class would go to some location [say the Market] and then pick their first command [something like “Find someone wearing a hat.”]. Once such a person was found, the group would pick a second command [something like “Follow the person until they go indoors.”]. Once the group was at the spot where the hat-wearing person went indoors, they would pick the last command [something like “Identify all the trees you can see.” or “Ask the first person that walks by about the history of this spot.”] The group’s observations where recorded in a wiki and they became entries in an atlas building exercise: the first two commands randomizing location and the last command identifying the info to be collected.


I have a class this term that is undertaking what I think is a similar activity, only this time in the blogosphere. They are finding content by choosing what sites to visit in a series of steps. The observation at each step is a word cloud of the site they have reached. I’m interested in how their “content map” compares to what various search engines reveal; the characteristics of the search paths; etc. This involves visualizing content and extracting content from word clouds.

Skip to toolbar