Mike, (01)
There is no reason why they have to be disjoint: (02)
> I think Matthew, MichaelG and CHrisM are all right.
> You need axioms when you get into the details.
> You also want a high level diagram the conveys the main
> points, w/o gory details. (03)
Many people are quite happy with UML-style diagrams, and there
is no reason why formal translations from UML to CL can't be
done. The reason why such formal translations haven't been
done is that "customers haven't asked for them". (04)
The solution is to develop tools that translate UML diagrams
into CLIF and use that as the de facto standard for what UML
means. To supplement the diagrams, use a version of controlled
English, such as CLCE or ACE. (05)
The CL model theory (or the subset called LBase) is already
the standard for OWL and RDF. If we adopt that for the other
tools, we automatically have a universal standard that lets
people use any pet notation they prefer. (06)
That's the proper way to develop interoperable standards:
start with the semantics, not the syntax. (07)
John (08)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uos-convene/
To Post: mailto:uos-convene@xxxxxxxxxxxxxxxx
Community Portal: http://ontolog.cim3.net/
Shared Files: http://ontolog.cim3.net/file/work/UpperOntologySummit/uos-convene/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UpperOntologySummit (09)
|