RE: Documents, Cars, Hills, and Valleys

 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Danny,

Contextual search is a simple enough idea: www.teoma.com .
Usability and IR people have been suggesting it for quite a few
years now as an augment to people using computers. Somebody smart
should figure out a similar hack for URIs in RDF, because there's
no chance people will use them in an unambiguous way as names
across the web. Using a thesaurus, or probabilities, comes to mind.
 
If the downstream value of RDF is in merging information, as some
of the people working on it for a few years now have suggested,
then I claim disambiguation as a requirement for the semantic web.
Assuming a URI identifies one resource is not a weak assumption;
assuming we'll all agree which resource is very weak.

Perhaps the web in place today isn't expressive enough to negotiate
disambiguation; we may have a bootstrapping problem.

> Ok, if I'm talking about the size of http://www.microsoft.com 
> am I talking about the number of employees or the number of 
> characters on their home page? 

My common sense says you're talking about the home page. After all,
you're using a http: scheme URI. Unfortunately common senses aren't
always interoperable. If one wants to go ahead and use http: scheme
URIs as proper names, ie the represented resource is a company
instead of home page, they could at least attempt to provide a
protocol to aid disambiguation. Be clear.

Bill de hÓra


-----BEGIN PGP SIGNATURE-----
Version: PGP 7.0.4

iQA/AwUBPMaX1eaWiFwg2CH4EQIYMQCfalHzkZe4whIX/wkdCsusS7SFeRkAoLrS
yBUbDpm1A2p4dCDhc2oH4WEx
=FqHc
-----END PGP SIGNATURE-----