On Thu, January 23, 2014 12:42, John McClure wrote:
> ...
> rdfs:subPropertyOf is interesting. It contains a preposition -- of --
> that we agree upon. Semantically it is a subclassing mechanism but one
> specific to properties, (01)
This is an interesting use of the term "subclass". I take it that you see
one
property as a subclass of another property, if its extent (i.e., the set
of valid sentences possible with that predicate) is a subset of the extent
of the second property. (02)
Extending this, you view a property as a class of sentences. (03)
> so there's no difference between
> rdfs:subPropertyOf and rdfs:subClassOf, with that exception of
> *context*. (04)
The difference is with argument type, with rdfs:subClassOf not accepting
the type of class which you consider a predicate to be (or at least not
accepting the name of a predicate as representing such a class). (05)
> Properties may only be a subclass of other properties and
> Classes may only be a subclass of other classes. And really, what is a
> subclass other than a subdivision? (06)
A subclass is a subdivision of a class. A subdivision of a physical
object is not a subclass. Nor is a subdivision of an organization, a
piece of music, a magazine article, or a game of cricket. (07)
> Accordingly, it could have been rdfs:subdivisionOf -- that would
> certainly be clearer to the average Joe, (08)
I don't think i've met your average Joe. I would guess that such a
general term would not make an average English speaker understand a
subproperty, while the term subProperyOf would have more of a chance to. (09)
> and it would have been
> reusuable in other *contexts* -- the operative principle of *reuse*. (010)
Reusable in other contexts means reusable with the same meaning when
applied to different sets of data -- not reusable with an analogous
meaning when applied to different types of arguments. (011)
> But instead, we get two oracular properties, one a completely *made-up*
> word (subProperty), (012)
A property is *not* a word, but it has a name. I prefer a system of
nomenclature in which descriptive names are not pre-existing words, but
are made from them. This leads to users being less likely to assume they
know the meaning without reading the definitions. (013)
> rather than community consensus about a more abstract concept. (014)
More abstract concepts are useful when rules are applicable at that level
and don't need to be restated with a more specific meaning at the more
specific level. When the more specific concepts have their own rules that
don't apply to the more general concept, it is better to use the more
specific concepts. (015)
> I'll address the generic nature of "is:of" later/jmc (016)
I agree with Krzystof below that the name "is:of" seems so generic as to
be almost devoid of meaning. What specific rules can be stated using
is:of that have no need to be further specified at a less generic level? (017)
-- doug foxvog (018)
> On 1/23/2014 9:21 AM, Krzysztof Janowicz wrote:
>> I guess you are confusing namespaces with linguistic expressions.
>> "is:about" would require an is namespace. How would id possibly help
>> to rename a human understandable label such as rdfs:subPropertyOf into
>> an overly generic is:of? is of what? is of blue color? One way to
>> address your issues is to join the W3C RDF working group and
>> contribute to their work to jointly improve RDF. Best, Krzysztof On
>> 01/23/2014 08:55 AM, John McClure wrote:
>
>
> _________________________________________________________________
> Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
> Subscribe/Config:
> http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
> Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
> Community Files: http://ontolog.cim3.net/file/work/OntologySummit2014/
> Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2014
> Community Portal: http://ontolog.cim3.net/wiki/
> (019)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2014/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2014
Community Portal: http://ontolog.cim3.net/wiki/ (020)
|