] [ Color me skeptical. For one thing, most of your data
] doesn't have any useful metadata, unless you put it there, and
] furthermore, google's not really the best thing to use when
] trying to determine the content of a document. It'll match
] keywords with other documents on the web, and it's reasonably
] smart about it, but i don't see a way to make it take my
] document and perform a semantic analysis on it. There's a lot
] of work going on in this area, but i don't think anything's
] broken past the basic experimentation level. I'd like to be
] proven wrong. -k]

http://www.nat.org/dashboard/

RE: Jesse Ruderman: Experience Google's new look