>> 1. Maybe the most important question is (and probably only for my
>information and understanding): why do we base the vocabulary on VIM?
> VIM is the International Vocabulary for Measurement, a publication of
> BIPM -- the people who maintain the international standards for units of
> measure. http://www.bipm.org/en/publications/guides/vim.html
><http://www.bipm.org/en/publications/guides/vim.html>
> It seemed like adopting their vocabulary would be good, and we have
> tried to stay close to that.
Agreed. (01)
>> (I'm very interested in studying VIM, but couldn't find the document. Could
>someone please send me a link?) There are also other documents that we could
>base our vocabulary on.
>>
> Yes. For example, we could use the model in DOLCE or Cyc or SUMO or
> some other published upper ontology that provides formal axioms for the
> quantity and measurement concepts. We have a work item that says we
> should look at these.
I was thinking more about ISO standards, or documents like the SUNAMCO
document. But please understand me correctly: I fully agree (and see now after
studying) that VIM is a very important document! (02)
>> 2. But then I ignore VIM, and repeat myself: couldn't we have:
>>
>> "length of my table" ---"member of" ---> "length quantity"
>> "length quantity" ---"subclass of" ---> "quantity"
>>
>> This is in accordance with some standard documents I know.
>>
> And it is in accordance with the model that David has produced. But he
> uses "instance of", instead of "member of" ("instance" is an
> "intensional" characterization; "member" is an "extensional"
> characterization; although probably only an ontology purist would care.)
It's great that it is in accordance with David's model. (As to the terms
"instance of" or "member of", they're both OK to me. I have never thought about
intenstional and extensional characterizations - seems like an interesting
topic. But indeed, we can probably better skip it now.) (03)
>>
>> 3. Also, according to these documents, e.g. "1.3 metre" is not a quantity,
>but a measure.
> Agreed. VIM calls it a "quantity value", which is an "expression of the
> magnitude of a quantity". VIM uses "measure" to talk about a process
> for determining a quantity value and for the results of that process.
Agreed. I'll come back to quantity values further below. (04)
>> (A quantity is like a variable, it is a reference to a value, like "1.3
>metre".)
>>
> We are careful to distinguish the class "quantity", whose instances are
> particular quantities, like "the length of your table", from the class
> "quantity magnitude", whose instances are the abstractions -- amounts of
> length stuff. So if you have two chairs next to your table and they are
> identical in size and shape, "the height of the first chair" is a
> (particular) quantity, and "the height of the second chair" is a
> (particular) quantity, and they are _different_ instances of "quantity",
> but they have the same "magnitude". That is, there is one amount of
> 'length' that they both have.
Agreed. (05)
> "1.3 metre" is a quantity value, which is an expression/name for a
> "quantity magnitude". If your chairs are 1.3 metres high, then "1.3
> metres" is a name for the magnitude of the height of each chair. And of
> course, the same magnitude can have other names, such as "51 inches".
> "51 inches" and "1.3 metres" are two different quantity values, but they
> express the same "magnitude". That is the model we are proposing.
Agreed. (06)
> But then there are two (or more) different classifiers that are called
> "length". length-1 is a subclass of "(particular) quantity". The
> length of your table is a length-1, the height of the second chair is a
> length-1. But the magnitude of the height of the second chair is not a
> length-1. length-2 is a subclass of 'quantity magnitude'. The
> magnitude of the height of the second chair, which is the same as the
> magnitude of the height of the first chair, is a length-2. So we must
> choose one of these to be what we mean by "length". VIM is very clear
> that it means length-1. But then we must not say that "1.3 metres
> expresses a length"; we must say (as VIM does) that "1.3 metres
> expresses the magnitude of a length".
If I understand everything correctly, agreed. Maybe things would become clearer
if we used a different term for "quantity magnitude" (or "quantity value"). I
would propose the term "measure". In the terms "quantity magnitude" and
"quantity value", I think, "magnitude" and "value" represent aspects (or
properties if you wish) of "quantity", rather than that they or the full terms
"quantity magnitude" and "quantity value" represent an independent construct,
i.e., the combination of a "number and reference together expressing the
magnitude of a quantity" (VIM (1.19)). The magnitude (or value) of a quantity,
then, would be a measure (I mean, the range of the property "magnitude" or
"value" would be "measure" (domain: "quantity"). (07)
> I personally think this usage is confusing for everyone. I would prefer
> that we use "particular quantity" to be the class of things like the
> height of the second chair, and use "quantity" to be what VIM calls the
> "magnitude". But, on the other hand, you can't ever measure a
> "magnitude"; you can only measure a particular quantity -- the height of
> the second chair. The "magnitude" itself is not a physical phenomenon;
> it is an abstraction. So it is very important that we agree on what
> definition we choose for each term we use. Otherwise, everyone is confused.
There are many ways to define all concepts, but I think the term "quantity" is
usually used as defined in VIM (1.1). I also base this on other standardization
documents. This is different with the term "measure", this is usually not so
explicitly defined. Alternative names do exist, such as, particularly,
"magnitude". See further above. (08)
> > Again, hopefully this discussion is appreciated; it is meant to be
>constructive, with the aim of working together to achieve a high-quality
>vocabulary.
> Me, too. But we keep going around these same issues, because people
> don't understand the definitions.
It's difficult. (09)
> -Ed
Hajo (010)
<<winmail.dat>>
_________________________________________________________________
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 (01)
|