Cart
[Home] [By Thread] [By Date] [Recent Entries]
Just map the vector model vectors to the URIs and replicate them. Not a very hard problem as long as the index types can be published. Basically it's the semantic web simplified by taking the schemas and giving them a vector, taking the same from the unstructured texts and mapping them. The real fun begins in a multi-variate document such as an aggregate where the sets of the single namespace become a term in the higher level space but that is all aggregation is anyway. More fun comes when you begin to treat non-text datasources such as audio and video as if they were texts and impose namespaces onto these and induce intent. len From: Michael Champion [mailto:michaelc.champion@g...] Thanks, I was hoping someone would start a thread about that! I have found a link to (perhaps some earlier version of) the presentation itself http://www.webratio.com/images/20050408Bosworth.pps I'm still digesting all this ... certainly my first reaction was that somehow putting web services and XQuery in the same category, and setting up RSS 2 as their antithesis, is stretching my mind in places it doesn't want to stretch :-)
|
