RE: web proper names

Hi Uche,

Uche Ogbuji writes:
 > On Sun, 2004-09-26 at 02:04, Phil Dawes wrote:
 > > Uche Ogbuji writes:
 > >  > > Using an IFP - e.g. foaf:mbox. 
 > >  > 
 > >  > Hmm.  My problem is that in the real, muddy world, I think
 > >  > machine-automated merging is going to run into many more ontological
 > >  > problems than you can throw a few OWL gimmicks at.
 > >  >
 > >  > Just ditch the idea: it's not in the least bit practical.  *People* will
 > >  > always be needed to sort out context, Semantic Web or no.  I think that
 > >  > once one accepts that, all these semiotic contrivances become easily
 > >  > seen for the nonsense they are.
 > >  > 
 > > 
 > > Actually I've been finding it very practical (at least working in an
 > > intranet environment). Of course you're right - people do need to sort
 > > out context, but they also need to be able to do this on a large scale.
 > 
 > "intranet environment" is a closed system.  For some reason you snipped
 > the part where I said that it *is* perfectly practical in closed systems
 > where the symbols can be controlled by management and policy.
 > 

Umm.. Apologies if you think I mis-quoted you, but I can't find any
reference to you saying that in the email I replied to.

Never mind. The point I was trying to make was that, yes, people are
required to sort out context, but that once they've decided that they
are able to use a set of symbols to mean something (within an
acceptable margin of error in the outcome), OWL gimmicks like IFP
etc.. are very useful in smushing/processing the data.

 > "intranet environment" is not "the real, muddy world", which is where I
 > said machine merging magic is impractical (because there is no centrally
 > controlled grounding of symbols).

But machine merging magic is driven by people based on context. I'd be
surprised if anybody thought otherwise.

Also note that even in a closed system with policy etc.. symbols still
squew. Even the owner of a symbol may use it to mean different things
in different contexts.

(e.g. http://databaseteam/Server meaning 'computer server' vs
'computer server that the database team manage' depending on context).

Cheers,

Phil

Received on Friday, 1 October 2004 08:43:51 UTC