RE: Registrars of terminology

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Certainly, and today there is only time for a 
superficial look.  But that goes to the issue that Passin raised in 
the Linda example.  How do you know how deep to go 
based on a pattern match rule?  Would this be one 
example of how one might want a Topic map such that 
the single term identifies ALL of the named items 
in the relationship?  Terse is a problem; but not 
having wide enough scope is another problem.  Two 
systems trying to negotiate based on the descriptions 
need rules for ensuring they are negotiating the 
same things.  Tough problem unless one uses authoritative 
assertion to declare the source as correct, i.e., the 
named record of authority that governs the negotiation 
process.

Identifying the right expert is the hardest problem 
after identifying the right negotiator.  So much 
of our time is spent filtering the encoding bias.

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Uche Ogbuji [mailto:uche.ogbuji@fourthought.com]

I still think it's somewhat valuable for the prose descriptions.  Len
pointed out that many of the descriptions themselves are terse, but I've
found that they tend to fill up a bit if you look at several related terms
and piece together the description.