Re: Generated RDF conformant with good practise?

On 17/09/2004, at 1:41 PM, Thomas B. Passin wrote:

>
> Howard Katz wrote:
>>> ah, I misunderstood what a bib was, sorry, serves me right for 
>>> rushing
>>> it. It might clarify things to have typed Bib and Book objects 
>>> perhaps.
>> OK, I didn't make that clear. Sorry. Given that understanding then, 
>> I'd like
>> to re-ask my question to see if your answer still holds. Assuming we 
>> want
>> our RDF to represent a bibliography containing pointers to various 
>> books, do
>> you still feel the following is a reasonable way of modelling that
>> relationship (ignoring the question of typed vs untyped nodes)?
>> <?xml version="1.0" encoding="UTF-8"?>
>> <rdf:RDF xmlns:bibterm="http://www.book-stuff.com/terms/"
>>          xmlns:dc="http://purl.org/dc/elements/1.0/"
>>          xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
>>     <bibterm:Bib>
>>          <bibterm:book rdf:parseType="Resource">
>>                  <bibterm:year>1994</bibterm:year>
>>                  <dc:title>TCP/IP Illustrated</dc:title>
>>          </bibterm:book>
>>          <bibterm:book rdf:parseType="Resource">
>>                  <bibterm:year>1992</bibterm:year>
>>                  <dc:title>Advanced Programming in the Unix
>> environment</dc:title>
>>          </bibterm:book>
>>      </bibterm:Bib>
>> </rdf:RDF>


This is a good example of something I have wondered about.  If in fact 
one wanted to create an RDF schema to describe such a structure, then 
it would be nice to be able to describe that the range of bibterm:book 
is any resource that has the properties bibterm:year and dc:title.  Is 
it possible to do this without making a class to represent that, or 
using the syntax and semantics of OWL?  I.e. is it possible in plain 
RDF/RDF-S to say that the properties of bibterm:year and bibterm:title 
are sufficient for a resource to be a valid instance?  if not, then 
perhaps it is worth considering a non-anoymous resource for this 
relationship so that one can still define the architecture of the 
bibterm:book property in RDF/RDF-S semantics.

cheers
Matt



>
> It looks good.  Let me translate this into words, and you tell us if 
> it captures what you intend.
>
> "There is a thing of type "bibterm:Bib", which bears a "bibterm:book" 
> relationship with something that has a "bibterm:year" property of 
> "1994" and a "dc:title" property having the value "TCP/IP"; ..."
>
> If that's what you meant, your rdf looks like the way I would prefer 
> to write it myself.
>
> Now you haven't said what type of anonymous thing has the bibterm:year 
> and dc:title properties, but maybe you don't care.  Or maybe you have 
> a few OWL statements that say the those two properties have a domain 
> of "bibterm:Book".  Then an OWL-aware reasoner might be able to infer 
> that those anonymous things are actually Books.  Or maybe not, if some 
> other property could also have those properties for its domain.
>
> Notice that by the judicious use of rdf:parseType (and maybe adding a 
> default namespace), you can sometimes convert  "ordinary" xml to legal 
> and appropriate rdf... it may take little or nothing more.  Say what 
> you will about rdf/xml syntax, I think this ability is a big plus for 
> it.
>
> Cheers,
>
> Tom P
>
> -- 
> Thomas B. Passin
> Explorer's Guide to the Semantic Web (Manning Books)
> http://www.manning.com/catalog/view.php?book=passin

Received on Monday, 20 September 2004 00:21:29 UTC