A Response to John Sowa, renamed from the thread "Are there primitive
concepts":
[JS] > If you want people to pay more attention to the COSMO ontology, you
> might show how COSMO terms can improve the Schema.org definitions.
>
I am willing to work with anyone who is implementing an ontology in some
practical application, provided that we can find a way to test proposed
improvements and use some objective metric to decide whether the application
does in fact work better. (01)
I took a look at the OWL version of the schema.org ontology, and although
there are reasonable (and sketchy) structures within the hierachy, there
appear to be significant problems from a logical perspective. Just one
example: (02)
There is a class 'CreativeWork' ("The most generic kind of creative
work, including books, movies, photographs, software programs, etc.") with a
subtype (via "MediaObject") of 'AudioObject' (An audio file.). OK so far.
*BUT* there is also a relation 'audio' which relates 'CreativeWork' as
domain to 'AudioObject' as range.
Since the Class 'AudioObject' is already a subclass of 'CreativeWork' ,
this relation does not function as a typical relation should, to relate two
or more entities in a manner to add some meaning. Instances of the class
'AudioObject' are already thus identified as audio objects, and do not need
the 'audio' relation to add more information. It's hard to see why that
relation would be at all useful.
Perhaps the ontologist had in mind that a 'CreativeWork' would be an
abstract class and 'AudioObject' is a physical object that is a
representation of an 'AudioObject' that can be played on an audio device.
But that is not what the ontology says. In fact, I can't find any
distinction between a 'CreativeWork' (which, if it is to be a named
individual such as "Star Wars" must be abstract with multiple physical
embodiments) and the physical embodiments themselves, such as tapes, DVDs,
books, still images, movie film strips, or computer files. If there is no
distinction, then there is no obvious way to relate the many different
physical exemplars of a creative work to each other. (03)
I will be happy to suggest alternatives that make more sense (to me)
logically, but that is only one of what may be many issues. (04)
I will be happy to discuss such issues with the developers of schema.org,
**or any other ontology**, but unless someone in the developing group
actually suggests some specific way I can be of help, and has an application
to provide a metric of utility, I would not take the initiative, but should
continue working on the COSMO itself for the immediate future. There are
still refinements that need to be made before I would want to use it in a
complex application like Natural Language, one of my primary goals. COSMO
as is may be adequate for less complex applications. (05)
Pat (06)
Patrick Cassidy
MICRA Inc.
cassidy@xxxxxxxxx
1-908-561-3416 (07)
>-----Original Message-----
>From: ontology-summit-bounces@xxxxxxxxxxxxxxxx [mailto:ontology-
>summit-bounces@xxxxxxxxxxxxxxxx] On Behalf Of John F Sowa
>Sent: Tuesday, February 11, 2014 11:17 PM
>To: ontology-summit@xxxxxxxxxxxxxxxx
>Subject: Re: [ontology-summit] Are there primitive concepts?
>
>Pat and Andrea,
>
>PC
>> I think we can agree that there will be newly discovered aspects of
>> reality that change our perceptions of what now seem to be primitive
>> concepts. So the inventory of "primitives" may change over time.
>
>That depends on what you mean by change. If you select something like the
>Longman's set of defining terms, then they can be relatively stable
>-- but only because their meaning is very loosely defined.
>
>Andrea stated that point quite clearly:
>
>AW
>> I was advocating under-specifying related (but not fundamental)
>> concepts in your modules (such as the Location concept in the Person
>> ontology example). Then combine modules that "complete" the
>> under-specified concepts - where the modules that you include are
>> consistent with your use cases and micro-theories ...
>
>The Schema.org terms are a large and growing set of useful but
>underspecified terms. Right now, those terms are defined by English texts
>that resemble OWL comments more than formal definitions.
>
>If you want people to pay more attention to the COSMO ontology, you you
>might show how COSMO terms can improve the Schema.org definitions.
>
>John
>
>__________________________________________________________
>_______
>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/ (08)
_________________________________________________________________
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/ (09)
|