At 3:53 PM -0700 7/24/08, Olasov, Ben wrote:
> From: [EMAIL PROTECTED] on behalf of Dan Russler
Hi Samson,
If "denote" = "describe" in your sentence, then I withdraw my objection.
But "denote" /= "describe". These two words
were intended to be, and are, used differently.
Consider the sentence, "The matrix P contains
the entries pij, denoting the probability that
for a speaker object i is associated with sound
j.". The word "denoting" functions here in the
sense of "marking" or "signifying". The word
"describing" has no equivalent sense that would
enable it to be substituted for "denoting" in
this sentence.
Indeed. And leaving ordinary English aside for a
moment, the formal meanings of 'denote' and
'describe' as used when discussing semantics of
formalized descriptions in logics and cognate
notations are not the same. Names denote things,
and descriptions - typically, extended
collections of sentences - describe them. They do
so by virtue of the names in the sentences
denoting the things being described, but the two
notions are distinct. In particular, denotation
does not require description. If I say, pointing
into the sky, "Look at that!" then my word "that"
denotes something I am pointing to. It does not,
however, describe it. Indeed, the thing denoted
may, like the Golux's hat, be indescribable.
I think we need to be very careful about
assumptions of denotational equivalence,
especially when the words under consideration
are the very ones that make the idea of
denotational equivalence possible and useful in
the first place.
I agree.
Pat Hayes
My concern is that the term "class" as used in UML doesn't seem to mean
the same thing as you are describing for a class in OWL. For instance, I
don't see the same concept of "resource" in the definition of class in UML.
A UML static diagram is just a symbolic method of displaying a set of
related assertions, i.e. attribute assertions, association assertions,
and state transition (behavior) assertions. The semantic interpretation
of what the class means comes not only from the text definition of the
"class symbol," but also from inferences made from the entire network of
attributes, associations, and state transitions.
To infer from a UML class more than is implied by the the above
statement is incorrect.
If we can agree on that, then we can better evaluate the equivalent and
non-equivalent semantics of OWL and the other methods for organizing
sets of assertions.
Dan
Samson Tu wrote:
>
>
> Dan Russler wrote:
>
>> Hi Samson,
>>
>> We are getting closer.
>>
>> 1) In the reference you site..."A class is the descriptor for a set"...
>>
>> 2) Earlier, you stated that "semantics of a class as denoting a set
>> of instances."
>>
>> I believe these two statements represent the "apples" and "oranges"
>> you referenced:
>>
>> Statement 1) is the traditional "a class describes the attributes and
>> associations for a concept that are common to a set of instances."
>> Statement 2) is better described by your population example.
>>
>> I wasn't objecting to 1) . I was objecting to your seeming to
>> confusie the 2) with 1).
>>
>> However, if you claim that "denote" means the same thing as
>> "describe," then I would agree with you instead of objecting to your
>> assertion.
>>
>> To be a little clearer....The definitions in a set of dictionaries
>> all "describe" the meaning of the word "farmer." However, the word
>> "farmer" in a dictionary does not "denote" the set of instances of
>> farmers in the world. Same with a UML class titled "farmer."
> >
> Dan,
>
> Yes, I am claiming that "denote" means the same thing as "describe" in
> my intended usage of the English words.
>
> We are talking about the semantics of "class." The class "farmer" is
> not the same thing as the dictionary word "farmer." Some people say
> that UML is just a graphical notation without semantics because it
> does try to make its meaning of the word "class" very clear. In
> logic-based knowledge representation languages, the set-theoretic
> semantics of class is widely used.
>
> The OWL Reference[1] put it this way:
>
> Classes provide an abstraction mechanism for grouping resources with
> similar characteristics. Like RDF classes, every OWL class is
> associated with a set of individuals, called the class extension. The
> individuals in the class extension are called the instances of the class.
>
> OKBC [2] p. 6 put it even more baldly:
>
> A class is a set of entities. Each of the entities in a class is said
> to be an instance of the class.
>
> In logical term, a class is a unary predicate satisfied by all of its
> instances. Dictionary definitions of words are not involved.
>
> [1] http://www.w3.org/TR/owl-ref/#Class
> [2] http://www.ai.sri.com/~okbc/okbc-2-0-3.pdf
>
>--
>---------
>Samson Tu email: [EMAIL PROTECTED]
>Senior Research Scientist web: www.stanford.edu/~swt/
>Center for Biomedical Informatics Research phone: 1-650-725-3391
>Stanford University fax: 1-650-725-7944
Ben
--
---------------------------------------------------------------------
IHMC (850)434 8903 or (650)494 3973 home
40 South Alcaniz St. (850)202 4416 office
Pensacola (850)202 4440 fax
FL 32502 (850)291 0667 cell
http://www.ihmc.us/users/phayes phayesAT-SIGNihmc.us
http://www.flickr.com/pathayes/collections