Quoting "Riley, Jenn" <[log in to unmask]>:
> Hi Karen,
>
> If it helps, the place I was looking for this information, both in
> the Guidelines and SWAP in particular, was in the "Domain Model"
> section. This is the section where the "things" in scope of the AP
> are defined, so to me it makes sense that this is where any formal
> definitions (whether simply available or required) for those things
> should be referenced.
I believe that the domain model is independent of RDF declarations. So
maybe there needs to be a section that explains how one goes from the
domain model to RDF/DSP formalities.
A question that I have, due to my naive understanding of RDF, is
whether there is a correlation that can be made between "entities" in
the "E-R" sense, and classes in the RDF sense. I know that the FRBR
entities have been modeled as classes in various implementations. The
DSP document does not equate description sets with classes, and it
looks like you can define and name a description set without a
corresponding class definition.
Someone else needs to weigh in here -- I have only questions, no answers. :-)
kc
>
> Thanks,
> Jenn
>
>> -----Original Message-----
>> From: Karen Coyle [mailto:[log in to unmask]]
>> Sent: Thursday, April 08, 2010 9:41 AM
>> To: DCMI Architecture Forum; Riley, Jenn
>> Subject: Re: DC Application Profile - require formal properties but not
>> classes?
>>
>> Jenn, I had to go back and browse through the DCAP guidelines to see
>> where this might fit in... the examples in the guidelines were mainly
>> focused on re-use of properties that have already been declared
>> elsewhere (and attempt to stay very very simple even in that). In that
>> case, incorporation of the property in an AP would by its nature
>> include the class relationship, or at least I viewed it that way. If
>> you look at the DSP document, which goes into a great deal more detail
>> (which we were trying to avoid in the guidelines), at 6.6.2 it
>> includes a class membership constraint. This, however, is only present
>> for non-literal values, and I have no explanation for that.
>>
>> Perhaps Appendix C of the guidelines should make some mention of
>> classes, with a reference to W3C documentation for that level of
>> detail?
>>
>> kc
>>
>> Quoting "Riley, Jenn" <[log in to unmask]>:
>>
>> > Thanks, Andy, that info is extremely useful.
>> >
>> > Perhaps SWAP could link to those RDF declarations from their AP
>> > documentation? And the DCAP Guidelines could be a little clearer in
>> > this area in their next revision?
>> >
>> > Jenn
>> >
>> >> -----Original Message-----
>> >> From: DCMI Architecture Forum [mailto:DC-
>> [log in to unmask]]
>> >> On Behalf Of Andy Powell
>> >> Sent: Wednesday, April 07, 2010 10:08 AM
>> >> To: [log in to unmask]
>> >> Subject: Re: DC Application Profile - require formal properties but
>> not
>> >> classes?
>> >>
>> >> Irrespective of what the current DCAP guidelines say, my
>> understanding
>> >> is that DCAPs are expected to declare RDF classes for each of the
>> >> "things" in the domain model.
>> >>
>> >> Re: SWAP classes... see:
>> >>
>> >>
>> http://www.ukoln.ac.uk/repositories/digirep/index/Eprints_EntityType_Vo
>> >> cabulary_Encoding_Scheme
>> >>
>> >> I'm not sure that we ever made available RDF declarations of these
>> >> classes or not? It would certainly seem good practice to do so.
>> >>
>> >> Andy
>> >
>>
>>
>>
>> --
>> Karen Coyle
>> [log in to unmask] http://kcoyle.net
>> ph: 1-510-540-7596
>> begin_of_the_skype_highlighting 1-510-540-7596 end_of_the_skype_highlighting
>> m: 1-510-435-8234
>> begin_of_the_skype_highlighting 1-510-435-
>> 8234 end_of_the_skype_highlighting
>> Jenn,skype: kcoylenet
>>
>>
>
>
--
Karen Coyle
[log in to unmask] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet
|