It's automated so doesn't work too well (which it expects and warns you about) but is a nice idea.
It's an open source project via Google Code, which I suppose means it was someone's bright idea (actually, it was Ben Campbell's bright idea) and it's just sort of happened as a little job to boost the Trust's online exposure.
While I find the simple design, dreadful 1990s typeface and shameless theft of web 2.0 terminology (beta, newsfeed, tagcloud) endearing, the most fun I had was rummaging through the issues that the developers faced while creating the site - good ones include:
David Leigh (Guardian) confused with David Leigh (Mirror)
(Oops. Not one to get confused.)
There appear to be some problems with the Independent scraper.
Some countries appear in the tags even though they were filtered out (e.g. Britain, Iraq, Iran, Russia)
(Why filter out Iraq?)
(Also published on LEWIS 360)