Re: Linked Library Holdings/Items

On Fri, Oct 21, 2011 at 3:23 PM, Jakob Voss <jakob.voss@gbv.de> wrote:
>
> { ?X ov:commonManifestation ?Y }
>
> => # implies one of the following holds:
>
> { ?X a frbr:Manifestation . ?Y a frbr:Manifestation . ?X owl:sameAs ?Y }
> OR
> { ?X a frbr:Manifestation . ?Y a frbr:Item . ?Y frbr:exemplarOf ?X }
> OR
> { ?X a frbr:Item . ?Y a frbr:Manifestation . ?Y frbr:exemplar ?X }
> OR
> { ?X a frbr:Item; frbr:exemplarOf ?Z .
>  ?Y a frbr:Item; frbr:exemplarOf ?Z .
>  ?Z a frbr:Manifestation . }
>

No, Karen's interpretation is correct.

These properties exist because FRBR is *so* rigid.  If you have a
frbr:Manifestation, it cannot have a dct:creator (it could, I think,
have a rda:statementOfResponsibility at most).  If you have a
dct:creator, in theory, the resource has some properties of a
frbr:Work (I think dct:subject might also be at the Work level).

So, if you had a resource like:

<http://example.org/myBook>
  a <bibo:Book>
  <dct:title> My Example Book ;
  <dct:subject> <http://example.org/subjects/1> ;
  <dct:creator> <http://example.org/identities/1>;
  <bibo:isbn10> 123456789X ;
  <dct:language> <http://purl.org/NET/marccodes/languages/eng#language> .

you have references to a Work (via the creator and subjects),
Expression (via the language), and Manifestation (via the isbn10).

Since frbr:Work, frbr:Expression, and frbr:Manifestation are all
disjoint with each other, <http://example.org/myBook> cannot be typed
any of these things (since it contains properties from all of these).
ov:commonThing <otherThing> just means that <otherThing> shares a
common frbr:Entity as <http://example.org/myBook>, but it doesn't
entail anything (except the inverse, <http://example.org/myBook>
shares a common frbr:Entity as <otherThing>).

I think the key to understanding this is in thinking about how
unforgiving FRBR is ontologically.

Does that clear more up?

-Ross.

> I can think of "share the same Work/Expression/Manifestation", but what does
> "share the same Item" mean? I bet that "A and B share the same FOO" must be
> expressable in RDF as
>
> <A> ?r _:x . <B> ?r _:x . _:x a <FOO> .
>
> So what about the relation ?r.
>
>> Hopefully, that will clear it up a little better.
>
> I'd better rename the properties to something like ov:sharesWorkWith or
> ov:hasSameWorkAs. The label "commonWork" looks more like "is common work of"
> which led me doubt.
>
> Jakob
>
> --
> Jakob Voß <jakob.voss@gbv.de>, skype: nichtich
> Verbundzentrale des GBV (VZG) / Common Library Network
> Platz der Goettinger Sieben 1, 37073 Göttingen, Germany
> +49 (0)551 39-10242, http://www.gbv.de
>
>

Received on Friday, 21 October 2011 23:03:01 UTC