xml-dev - Re: A Light Rant On Ontological Commitment

[Martin Bryan <mtbryan@sgml.u-net.com>]

Len Bullart wrote::
>
> To provide some thoughtMunchies for the first of the
> two questions I posed, here is an extract from
>
> "Conceptual Modeling for Distributed Ontology Environments"
> http://www.ontology.org/main/papers/iccs-dlm.html
>
> by Deborah L. McGuinness, Associate Director and Senior Research
Scientist,
> Stanford University, Knowledge Systems Laboratory

Great list Len.  Here are some I would suggest modification to:

> Specify mappings between multiple standard controlled vocabularies if
> multiple
> vocabularies are standard in the domain(s) of interest.

Specify mappings between multiple standard controlled vocabularies wherever
their domains of interest overlap. (This helps vocabularies to grow)

> Specify semantics of terms.

Provide multilingual mappings for terms and definitions

> Specify domains and ranges of roles (for example, the domain of
> "leader-of-country"
> is "person" and the range is "country").

Specify region and timeframe over which domains and roles apply. (What dates
will George Bush be leader-of-country? Which country? How does this differ
from being leader of Texas?)

> Specify active inferences to infer constraints between role values (for
> example,
> "shipping-weight" is greater or equal to the "actual-weight").

Domain dependent mappings are required to understand the relationship
between equivalences (US lbs and metric kg being a case in point, not to
bring up the subject of the speed of Mars landers!)

Martin Bryan
Chair, CEN/ISSS EC Workshop working group on Defining and Managing Semantics
and Datatypes
(Just to let you know where these thoughts are coming from!)