First, thank you, Doug, for actually answering the questions that Pat
Browne asked, at least for DOLCE.
No one seems to have answered for BFO. (01)
It seems to have escaped John's notice that Pat did not ask what
Aristotle thought a category is but rather what DOLCE and BFO _define_
it to be. This resulted in the usual Ontolog Forum diversion of the
discussion, which may be educational, but non-responsive. (02)
We must all recognize that 'ontology' very definitely takes the Humpty
Dumpty view that a word means what we say it means. You can only impose
your preferred ontology on that of another author where you find
coincidence of concepts or inadequate definition. (03)
This takes us to a favorite 'ad populum' mechanism for determining
appropriateness of terms and definitions. (04)
John F. Sowa wrote:
> Doug and Erick,
>
> DF
>
>> Unlike John, i do not mind the use of the term "subsumption".
>> His "subtype" or subclass are more specific terms; binary and higher-
>> order predicates can also be arranged in subsumption hierarchies although
>> not in subtype/subclass hierarchies.
>>
>
> My comment about 'subsumption' is just that it's a rare word that
> tends to frighten students. My recommendation is to use the term
> 'generalization' instead. From Google,
>
> Number of hits for 'subsumption' - 506,000
>
> Number of hits for 'generalization' - 13,200,000
>
> Number of hits for 'generalisation' - 2,420,000
> (05)
Now, the question is: How many of the hits on 'generalization' actually
intended it to be interpreted as the 'subsumption' relationship between
classifiers? (06)
'Generalization' is most commonly used among speakers of English to
refer to a statement that represents a statistical abstraction of
situations. A 'generalization' is (or may be) logically false, by dint
of admitting counterexamples, but it is true in some way of a
statistically dominant percentage of the 'particular' cases. Example:
"As a generalization, the Ontolog Forum traffic is not worth reading." :-) (07)
So, since Google represents a wide cross-section of users of English,
very few of whom really understand the concept of 'subsumption' among
classifications at all (a generalization, of course), I would say that
the assumption that 'generalization' is more frequently used than
'subsumption' _for the concept in question_ is unwarranted. And even if
it prove to be true, I expect the margin to be less than 2-to-1,
possibly in direct proportion to the ratio of description logic papers
to UML papers. (08)
> Another advantage of the noun 'generalization' is that it has
> the inverse 'specialization', and both nouns have common verbs
> and adjectives like 'general', 'more general', 'generalize',
> 'specialize', 'specialized', 'more specialized', etc.
> (09)
And thus are used to mean many different things. In general, there is
no relationship between 'general' and 'generalization' (of the
subsumption kind). 'Specialization', however, almost always has the
right general understanding. (010)
I fully agree that 'generalization' and 'specialization' are terms we
may want to use in our ontologies, and they should be well understood by
knowledge engineers of all kinds. But that only reflects the fact that
others have used the terms for this purpose in related disciplines for
over 30 years. It is an accepted engineering terminology. The fact is,
however, that 'subsumption' is an accepted term of art in the
description logic field (and others), which is of similar vintage, and
whose only weakness is that it had a much smaller population until recently. (011)
-Ed (012)
P.S. I do expect John to tell us that Charles Peirce used both of these
terms... :-) (013)
> If we want more people to learn and use ontologies, we need to find
> or define terminology that is precise, easy to remember, and easy
> to read, write, speak, and teach.
>
> EA
>
>> could you give an example of what you mentioned:
>>
>
> JFS
>
>>> In different situations, the same method of classification will
>>> produce a different set. The class remains the same, but the sets
>>> have different members or elements
>>>
>
> Pick a term for any species: 'dog', 'dandelion', 'E. coli'.
>
> Pick any situation: New York City at 12 noon today, Planet Earth...
>
> The predicates is-dog(x), is-dandelion(x), or is-E-coli(x) are
> specified by some description of the species. When applied to
> different situations, they will determine different sets.
>
> New York City is a big place, and it would be difficult or
> impossible to enumerate all instances of dogs, dandelions,
> or E. coli. But we can be certain that the sets at different
> points in time are very different.
>
> If you like, you can choose a smaller situation, such as
> the living room of your house. It's much easier to observe
> the sets of dogs and dandelions in your living room. If you
> don't have a dog the set is empty, unless you have a visitor.
> If you do have a dog, the set keeps changing every time your
> pet goes in or out.
>
> John
>
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
> Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
> Shared Files: http://ontolog.cim3.net/file/
> Community Wiki: http://ontolog.cim3.net/wiki/
> To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
>
> (014)
--
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 Cel: +1 240-672-5800 (015)
"The opinions expressed above do not reflect consensus of NIST,
and have not been reviewed by any Government authority." (016)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (017)
|