W3C home > Mailing lists > Public > semantic-web@w3.org > August 2008

Re: Why do you want to do that?

From: Frank Manola <fmanola@acm.org>
Date: Thu, 14 Aug 2008 11:34:00 -0400
To: "Richard H. McCullough" <rhm@pioneerca.com>
Message-Id: <059BE1C4-012D-4C2B-8738-A1FACE905E70@acm.org>
Cc: "Michael Schneider" <schneid@fzi.de>, "SWIG" <semantic-web@w3.org>, "KR-language" <KR-language@YahooGroups.com>

Dick--

See below.

On Aug 13, 2008, at 9:36 PM, Richard H. McCullough wrote:

>
> Hi Michael
> See below.
>
> Dick McCullough
> Ayn Rand do speak od mKR done;
> mKE do enhance od Real Intelligence done;
> knowledge := man do identify od existent done;
> knowledge haspart proposition list;
> http://mKRmKE.org/
>
> ----- Original Message ----- From: "Michael Schneider"  
> <schneid@fzi.de>
> To: "Frank Manola" <fmanola@acm.org>
> Cc: "SWIG" <semantic-web@w3.org>; "Richard H. McCullough" <rhm@PioneerCA.com 
> >
> Sent: Wednesday, August 13, 2008 4:57 PM
> Subject: RE: Why do you want to do that?
>
>
>> Hi Frank!
>>
>> Frank Manola wrote:
>>
>>> There aren't any classes in RDF (per se);  in RDFS there are  
>>> classes,
>>> and they can be treated as individuals (which is where we came in,
>>> more or less).
>>
>> As a minor remark, I think it's easier to understand the situation of
>> classes in RDFS, if one states the above sentence the other way  
>> around: In
>> RDFS there are individuals (aka resources), and some of them can be  
>> treated
>> as classes, namely those which happen to have a class extension  
>> associated
>> with them. Analog, some individuals have a property extension  
>> associated
>> with them, and are therefore properties. An individual may even act  
>> as both,
>> a class *and* a property, if it has both a class extension and a  
>> property
>> extension associated.
>
> But in any case (as you say), all classes and
>> properties are individuals, which exist in the RDFS universe, i.e.  
>> the
>> domain of discourse.
> ***** This is an unnecessary "corruption" of the concept hierarchy.
> Once again, the simple alternative is to use sets.
> all class ismem ClassSet;
> all property ismem PropertySet;
> all individual ismem IndividualSet;
> Except for the last line above,
> x being a member of a set does NOT make x an individual.

This discussion seems to illustrate the meat of the matter.  Speaking  
generally, it seems to me that, rather than treating properties and  
classes as individuals being an unnecessary *corruption* of the  
concept hierarchy, *not* treating them as individuals is an  
unnecessary *simplification* of the concept hierarchy.  The key word  
that applies, though, is "sometimes".  The next line below seems to me  
to say that you want to be able to do that in some situations  
("contexts", if you must use that word :-)   ).  That suggests that  
your language/model must support that capability (that is, if you  
*ever* want to do that, the capability must be there).  No one is  
arguing that you need (or even want) to be able to do that in all  
situations (at least I'm not so-arguing).  That people can  
successfully use OWL-DL shows that this capability is not needed all  
the time.  But it's certainly useful to have that ability in some  
cases.  We couldn't even be having this discussion without the ability  
to treat classes and properties as individuals!  Recall that the  
original question was "Why do you want to do that?"  I think there  
have been several examples of why given (including the one you cite  
below).  Dick, your argument now seems to be that you *never* want to  
do that *in the same situation*.  Is that so?

--Frank

>
>
> In an appropriate context, you might view a property as
> an individual.  For example
>   John Doe has happy;
> But in other contexts, it seems better to view a property
> as a Class.  For example
>   John Doe has gender = male;
> These examples correspond to the hierarchy fragment:
>
> begin hierarchy example;
> Property;
> /    i:happy;
> /    gender;
> //        i:male;
> end hierarchy example;
>>
>>> That is, in RDFS a class is a resource (like
>>> everything else that can be referred to in RDFS), and resources  
>>> can be
>>> the subjects of triples.
>>>
>>> --Frank
>>
>> Cheers,
>> Michael
>>
>
>
>
Received on Thursday, 14 August 2008 15:34:40 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 21:45:24 GMT