Wednesday, July 30, 2003

Google and the Semantic Web. Mark Watson made me think a bit about this.  He describes writing a spider that goes looking for RDF, and was disappointed that there's so little out there.  The thing is, manually generating all that stuff is just pretty much not going to happen. On the other hand, we have the GoogleGod out there, with the web's biggest pile of data.  That data has already been sliced up by word count, which is...you guessed it...exactly one half of a Bayesian classification system.  Yup, the same kind used to fight spam.  What if we applied the same technique to the entire web, via Google's tech?  Pretty cool, I think.  You be able to construct any number of classifications of information.  Each of those classifications can be described with RDF.  It can even be automated.  Pass a web page through some kind of google-based application, and what spits out is tagged with RDF, ready for the semantic web.  Companies could even be formed around providing this service. Ah, late nights.  1:18:26 AM