Martin S. Weber wrote:
> Ed Barkmeyer wrote:
>> (...)
>> By comparison, Martin Weber said:
>>> This is a confusion that should be equivalently confusing modelled! [i.e.
>>> keep all the overloaded meanings of the term 'quantity' to stick to
>BIPM-parlese].
>> I could not disagree more! (But I assume Martin was joking.) ...
>
> Actually I wasn't joking. I'm a CS. A CS is a CS or a CS with humor. (01)
How unfortunate! My job title is the same, but then I am not my job
title. ;-) (02)
> IMO the concept of a quantity (by means of the BIPM) refers to multiple
>levels
> of abstractness trickling down to specific things. (03)
Well, OK, as long as you can characterize the levels, you have n
different concepts and some "one concept to bind them all". (04)
> Do we want to have three different "length"s though that all refer to
> different abstract levels, one being the base quantity of the ISQ SI; one
> being all measurable lengths including the equivalent quantities distance,
> height and so on; one being specific length of things ? (05)
The answer may be YES, if these are all distinct and useful concepts.
So let's see. (06)
The SI base quantity "length" is a 'kind of quantity', such that all
things that are instances of "length" are comparable. It follows that
that same "length" is a subclass of the general class "particular
quantity" (so that it can have instances) and its instances must be all
the specific lengths of specific things. (07)
Thus far I have one "length" concept, which is an instance of 'kind of
quantity' and a subtype of 'particular quantity'. And we have many
instances of this concept "length", which are specific physical
phenomena to which the "length" concept applies. (08)
The other terms Martin introduces are "the equivalent quantities
distance, height and so on". I would describe these as either:
- subtypes of length, denoting subsets of the physical phenomena that
are distinguished by some further characteristic in the nature of their
measurement, or
- roles of length magnitudes or length quantity values in statements
about particular quantities that are instances of 'length'. (09)
So I see one class called "length" and its instances, and its membership
in a class ('kind of quantity') whose instances are classes. I also see
a set of additional terms that denote classifiers that are somehow
related to length. I guess that is "multiple abstract levels". And
yes, we do want the "length" concept and its relationships. I'm not
sure about the additional terms. (010)
> There is something bitter to bite into at either end and IMO the established
> bitter is preferable, isn't it? Adding another set of definitions that then
> are satisfying the ontological community but confusing the physicists,
> mathematicians (who happily deal with ambiguities like these) and computer
> scientists (whose daily bread often is structural recursions like these)...
>is
> that better? (011)
Yes, if the objective is to produce a rendition of our knowledge that
can be used by a reasoning engine to make correct inferences about
measurements. (012)
The above model is easy to capture in OWL and many other k-e languages.
It involves knowledge engineering concepts, not programming concepts,
not database concepts, not XML concepts, and not mathematics concepts
and not physics concepts. But it _captures_ physics concepts. (013)
The question is: Does it capture the INTENT of the physicists in a
USEFUL way? They don't have to read it. It just has to produce results
consistent with their expectations. (014)
> The strength of all these standards is their domain of
> application, where at latest people from other fields have to get involved.
> Confusing them and breaking their language habits isn't particularly helpful
> [again IMO]. (015)
If their language is ambiguous, how do you support their bad habits and
get good results? ("'Forward the light brigade! Charge for the guns!'
he said. Was there a man dismayed?" Unlike the 600, ours _is_ to reason
why?) (016)
IMO, (I repeat) the audience for this standard is knowledge engineers
who need to make domain-specific models that deal with measurement
units. It is NOT the domain experts. Yes, they provide the input. Yes,
we have to verify the model with them. That is not the same as making
them read it. (017)
-Ed (018)
--
Edward J. Barkmeyer Email: edbark@xxxxxxxx
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528
Gaithersburg, MD 20899-8263 FAX: +1 301-975-4694 (019)
"The opinions expressed above do not reflect consensus of NIST,
and have not been reviewed by any Government authority." (020)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uom-ontology-std/
Subscribe: mailto:uom-ontology-std-join@xxxxxxxxxxxxxxxx
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/uom-ontology-std/
Shared Files: http://ontolog.cim3.net/file/work/UoM/
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UoM_Ontology_Standard (021)
|