Re: (Round 2) Proposed Extensions to OWL

[Roger L. Costello]

To Recap:

1. To define a units-of-measure value requires a 3-stripping-layer
design.
2. A Units Conversion Ontology defines the mappings from one
unit-of-measure to another.
3. An ontology the uses units-of-measure (e.g., River Ontology) uses the
Unit class in the Units Conversion Ontology.

That's it.  What do you think?  /Roger

[Tom P]

Nice work, Roger.  Looks right to me.  I'm not all that sure about the
specific transform example you gave because I am unsure about how to connect
the formal parameters  to the slots in the MATHML formula to the actual
values in instance measurements.  But I am sure that can all get worked out.
The principle is right.

The key point now is this -  can this scheme successfully represent every
one of the "unit tests" we proposed?  Obviously most of them would pass, and
almost certainly all would except for those relating to the actual design of
the transform.

Also, I think the scheme is clean and well grounded in the real world of
values and measurements, and the complexities are isolated to places that
make sense (like the details of the transforms).

The one open area, at least for me, is how to abstract the equivalence
statements from specific instances to become ontological constraints.  I am
not well-versed enough in these things to be able to just whip up the
answer.  I hope someone else is.

Cheers,

Tom P

Received on Thursday, 10 July 2003 00:45:38 UTC