26.01.2017, 09:25, [log in to unmask] kirjoitti:
> What definition of Work will give the best results for the community
> that actually needs to use this data to actually do something useful?
> For example:
> · Cataloging. Is it really to the benefit of catalogers that the
> Work be split (or not split)?
> · Discovery
> · Work information as used by collection management or acquisition
> · The rest of the world
I'm currently working on converting the bibliographic records of the
National Library of Finland into BIBFRAME and Schema.org [1,2]. In the
process I've been thinking about the different ways Works are defined in
FRBR and BIBFRAME. (Schema.org with the bibliographic extensions can be
used to express Works the same way as BIBFRAME)
My current opinion on this is that the BIBFRAME model, where bf:Work
corresponds to the combination of FRBR Work and Expression, generally
makes things simpler for conversion - you don't have to worry so much
about whether something is merely a FRBR Expression of the same FRBR
Work, or a separate (FRBR) Work. It also simplifies the data model since
there are fewer entities in the chain.
However, some things end up being duplicated or awkward. For example,
subject description and classification is generally something that
should be applied at the FRBR Work level (though there are exceptions in
the actual records, for example I've seen "translated literature" as a
650 subject for a translated book). So if you have records both of the
original work and its translation, the subject and classification
information may have to be pooled from both and replicated one way or
both ways between the two bf:Work entities.
Of course this is nowhere as bad as with the original MARC records,
where all subject description fields are duplicated across all the
records of a single Work.
D.Sc. (Tech), Information Systems Specialist
National Library of Finland
P.O. Box 26 (Kaikukatu 4)
00014 HELSINGIN YLIOPISTO
Tel. +358 50 3199529
[log in to unmask]