Confused over use of DC vocabulary

[Carl Lagoze]

Hi,

Excuse me if this is been asked before or if I'm missing something obvious, I am new to this group.

I have been looking at bibo in an attempt to integrate with our work on OAI ORE and related work in chemistry modeling. I have looked at the latest version of the vocabulary distributed from http://code.google.com/p/bibotools/source/browse/bibo-ontology/tags/1.0/bibo.n3 and notice that there is a redefinition of the Dublin core terms as owl properties and classes. I understand that there is no established owl expression of Dublin core, so this is why you had to do this in your effort to use these vocabulary terms within bibo. However, I am confused that you then uniformly define all DC properties as owl object properties, for example:

<!-- http://purl.org/dc/terms/rights -->

    <owl:ObjectProperty rdf:about="&terms;rights">

        <skos:scopeNote xml:lang="en"

            >Used to describe rights related to a bibliographic resource.</skos:scopeNote>

    </owl:ObjectProperty>

and 

<!-- http://purl.org/dc/terms/title -->

    <owl:ObjectProperty rdf:about="&terms;title">

        <skos:scopeNote xml:lang="en"

            >Used to describe the title of a bibliographic resource</skos:scopeNote>

    </owl:ObjectProperty>

This appears wrong, because for these particular properties and many of the other DC elements, the value is explicitly a literal according to the Dublin core documentation, yet the object property designation demands that the object be a resource.

To make matters even more confusing, in your example serializations you use the same properties with objects that are indeed literals. For example:

<urn:isbn:23983498> a bibo:Book ;
    dc:creator <http://examples.net/contributors/2> ;
    dc:title "Book Title"@en ;
    dc:date "2000" ;
    dc:publisher <http://ex.net/agents/1> .

 is there something I'm missing here? Thanks for your answer.

Carl Lagoze

[Bruce D'Arcus]

On Thu, Jul 9, 2009 at 2:40 PM, Carl Lagoze<cla...@gmail.com> wrote:
> Hi,
> Excuse me if this is been asked before or if I'm missing something obvious,
> I am new to this group.

HI Carl,

You'll need to hear from Fred about the OWL stuff, but I'd just point
out that ....

> I have been looking at bibo in an attempt to integrate with our work on OAI
> ORE and related work in chemistry modeling. I have looked at the latest
> version of the vocabulary distributed from
> http://code.google.com/p/bibotools/source/browse/bibo-ontology/tags/1.0/bibo.n3

... the current version of the schema is 1.2. Would be better to
focus on that or the trunk version (though not sure it's any different
ATM).

Bruce

[Mark Diggory]

> HI Carl,
>
> You'll need to hear from Fred about the OWL stuff, but I'd just point
> out that ....
>
>> I have been looking at bibo in an attempt to integrate with our work on OAI
>> ORE and related work in chemistry modeling. I have looked at the latest
>> version of the vocabulary distributed from
>> http://code.google.com/p/bibotools/source/browse/bibo-ontology/tags/1.0/bibo.n3
>
> ... the current version of the schema is 1.2.  Would be better to
> focus on that or the trunk version (though not sure it's any different
> ATM).
>
> Bruce

Looking quickly, it is the same case here too
http://code.google.com/p/bibotools/source/browse/bibo-ontology/trunk/bibo.n3

--
Mark R. Diggory
@mire - http://www.atmire.com

[Carl Lagoze]

Yes, I noticed its the same in the new version. I'll wait for Fred's
input.

[Frederick Giasson]

Hi all,

It appears that this has been introduced by Protégé 4 at importation
time. The only thing I defined is everything related to BIBO; otherwise
they are all imports from protégé. Since it was making sense to import
all ontologies re-used in a BIBO context, it doesn't make sense if the
importation of these ontologies denature them. I think that for the next
version, I will remove the imports and only refer to them (as we did in
the past), and re-generate the documentation & sources.

Sorry about this,

Thanks,

Fred

[Bruce D'Arcus]

I wish the DC group itself would be more committal on a title being a literal.

Bruce