[open-bibliography] More verbs. Electronic 'Items' (Yes, another FRBR thread)
kcoyle at kcoyle.net
Thu Jul 8 18:33:07 BST 2010
Quoting Benjamin O'Steen <bosteen at gmail.com>:
> I was talking about realining or adding to the predicates when you take
> into consideration that some Expression-level predicates are arguably
> Work level relationships (or should be): eg translations and so on. For
> example, changing the range+domain of
> (hasATranslation) from Expression to perhaps a union of Expression and
> Work? or simply from E -> E to W -> W?
I'm still not clear on what you are trying to do, but perhaps this helps:
... in the FRBR document "is expression of" is a predicate that
relates an expression to a Work. Between each of the four Group 1
entities there appears to be a single relationship:
- is expression of (has expression) E <-> Work
- is manifestation of (has manifestation) M <-> E
- is item of (has item) I <-> M
As for translations, Work is an abstract, until it is expressed, so
only the Expression can be translated, not the Work. (That's the FRBR
logic, I'm not necessarily defending it.)
I do think it is problematic that relationships have been specified as
relating only between certain FRBR entities. In some cases I can see
the logic (as in above on translations), in other cases I cannot.
There's a lot about the implementation of FRBR that I think is far
Actually, it's worse than that. A predicate is a relationship between
properties, yet the FRBR predicates have been made specific to a
particular entity. It's not clear to me whether the predicate as
defined is subject + predicate or predicate + object.
Note that in the publishing of RDA (which uses a lot of FRBR elements)
we have tried to create a "general" level where properties and
predicates are free of the FRBR entities -- in particular the Group 1
entities -- so that communities with different interpretations of FRBR
can continue to operate.
> (predicate codified into RDF here:
> Do we have a list of relationships where the range and domain need to be
> loosened, based on previous discussions? Will we require more
> relationships to describe the Works properly?
> On Thu, 2010-07-08 at 06:41 -0700, Karen Coyle wrote:
>> Verbs can be found at:
>> Quoting Benjamin O'Steen <bosteen at gmail.com>:
>> > Two hopefully quick points:
>> > After the massive FRBR thread, I came away with the notion that Work and
>> > Expression are redundant levels, and that by redrawing the lines between
>> > the layers and adding in more verbs we could better serve the aim of
>> > FRBR'ising the records.
>> > So... where are the verbs? I've been moving house etc so haven't
>> > followed the threads as close as I'd like, but I do think this is a
>> > perfect opportunity to collaborate on a list of verbs that interrelates
>> > the classes that remain - the new 'Work', 'Manifestation' and 'Item'
>> > Shall we kick off an etherpad for this? or has someone already started
>> > collating predicates and I am just behind the times?
>> > --------------------
>> > Electronic Items
>> > Item doesn't make an awful lot of sense with regards electronics or
>> > born-digital items. I think that we might broaden the schema a little
>> > here to allow something like an OAI-ORE Aggregation here instead of an
>> > Item. Would it be preferable to subclass it as 'DigitalItem' or to use
>> > it unaltered?
>> > This would enable us to describe whatever granularity of digital item we
>> > want or are able to using RDF relationships.
>> > Ben
>> > PS forgive me if I have just summarised previous conversations - I am
>> > catching up!
>> > _______________________________________________
>> > open-bibliography mailing list
>> > open-bibliography at lists.okfn.org
>> > http://lists.okfn.org/mailman/listinfo/open-bibliography
> open-bibliography mailing list
> open-bibliography at lists.okfn.org
kcoyle at kcoyle.net http://kcoyle.net
More information about the open-bibliography