First of all, this is not my work. All credit should go to Tony Hirst from the Open University for posting an explanation of this to his blog. Brilliant stuff.
Tony is doing what we’ve been meaning to look at for some time, which is create ‘Learning Objects’ out of mashups. In Tony’s example, data is 1) taken from Wikipedia, 2) imported to a Google Spreadsheet, 3) output as a CSV file and 4) mashed up in Yahoo! Pipes.
The process of creating an object like the map below is so transparent that both teachers and students, with a specific outcome in mind, could achieve something like this and the results are very satisfying, as you can see.