From: Adam Pease (apease@ks.teknowledge.com)
Date: 03/10/03
Pete, That sounds like a lexicon issue, rather than an ontology issue. If a human user wants to use a particular term, that's an issue for the interface. A lot of problems occur when one tries to adjust an ontology, which must support inference, to conform to a human's linguistic conventions. We've found that it's straightforward to keep the two distinct, and have a lexicon file that maps to terms in an ontology. That also makes it easier to use the same ontology for users who speak different languages, as well as those who use different lingo. The issue of vagueness in language is related. A city is a region, which may have a centroid point that is represented on a map. A user may wish to be vague about that, but that's a language issue, not an ontology issue. Adam At 11:06 AM 3/10/2003 -0500, Pete Haglich wrote: >We have adopted that in our current spatial ontologies, using the term >"locus" because "location" had connotations to our customer. The notion >of "locus" has child subclasses of Point Locus, Linear Locus, Area Locus, >and Volume Locus. > >We find that the blurring of distinction between point and area is >sometimes useful when discussing things like cities. For some purposes >they are usefully modeled as points, in other cases, it is more >appropriate to reference the associated territory within the city limits. > > >On Friday, Mar 7, 2003, at 14:55 US/Eastern, Austin Tate wrote: > >>The idea being that a neutral word like location which does not restrict >>the meaning to be only one of a "point" or "area" or "volume" and leaves >>nature of the "location" open can be useful to allow relationships to be >>stated to give real meaning. >-- >Pete Haglich, ISX Corporation >Virginia Beach, VA >Mobile (757) 572-5913
This archive was generated by hypermail 2.1.4 : 03/10/03 EST