[open-bibliography] OKFN blog: Bibliographica, an Introduction

Mon May 24 00:26:58 UTC 2010

Quoting William Waites <william.waites at okfn.org>:

>
> Interesting. I wonder why they wouldn't just adopt the existing
> namespace.

Well, they *are* the developers of FRBR, so in a sense the FRBR  
namespace should be decided by IFLA, since they will be the ones  
maintaining the definitions. If you look at what is in the metadata  
registry, you will see that FRBR has evolved since the vocab project  
defined it, and it would be best to use the one that is the most up to  
date.

> Their URIs are also funny.
>
>     :WarAndPace a <http://iflastandards.info/ns/fr/frbr/frbrer/1003>
>
> regardless of namespace, it's not nice to give your classes numbers
> instead of names.

That's one of those religious questions :-). Some folks think that  
using names is not nice because it isn't language neutral. You'll find  
arguments no matter which method you use. For RDA properties, the RDA  
folks did decide that they wanted language-based URIs, and they've  
ended up with things like:

http://rdvocab.info/Elements/configurationOfPlaybackChannelsManifestation
http://rdvocab.info/Elements/otherDistinguishingCharacteristicOfTheExpression

Those make numbers look good.

>
> This would be most helpful. Have you read the bit in the documentation
> about using "lenses" with the Fresnel vocabulary (like "views") for
> visualisation? A lot of the work is going to be in creating those, and
> they're used both for editing and viewing.

No, but I will. And I'll think about it all, see what ideas I can come  
up with.

>
> That said, we run into a hard problem here. This is not intended to be a
> catalogue maintained by trained librarians but by researchers and
> scholars in other fields. How much do we require them to understand the
> FRBR conceptual model? For example, if they have just read Leopold
> Auer's "Violin Playing as I Teach it" and want to make an annotation to
> the effect that a particular chapter contains some good insight into the
> expression of Bach's Chaconne, do they make this annotation in reference
> to the Work, Expression, Manifestation?

... snip...

Given that even the developers of RDA can't seem to agree on what  
elements are valid for what FRBR levels, I think it would be best to  
skip FRBR altogether in the user view, and to let the data elements  
themselves carry the connection to the WEMI entities. There are not  
that many properties that can be associated with more than one level  
(at least not many that a simple catalog would use -- the majority of  
the multi-level properties are associated with both Manifestation and  
Item, and that is in support of the archival community's cataloging),  
and I think it's legitimate to choose to associate each property with  
a single FRBR level as a way to simplify the (overly complex) model.  
Subjects are associated only with the work in FRBR; language of text  
is associated with the expression; dates of publication and publishers  
with the Manifestation. FRBR would treat reviews or comments as  
separate works linked to the entity being reviewed.

So in response to Ross's comment:

> After all, if the only data you have is
manifestation level, it's not possible to (or necessary) to define the work
or expression.  That's the benefit of the open world model, somebody else
can fill in those parts you don't currently know.

... in fact, a Manifestation level record would have no creator and no  
subjects, also no language of text. FRBR, IMO, is a huge mess the way  
that it is actually written, and I am determined to find a way to make  
it more useful for real life applications. But the fact is that a  
"normal" bib record with author, title, publisher, subjects, is a  
mixture of WEM. There really is no such thing as a Manifestation  
record without WE. So that's why I feel like putting Manifestation as  
a type on the records is incorrect and could in the long wrong create  
more confusion (than FRBR already does on its own).

  On the other hand if someone wants
> to say that X's translation into Estonian of "100 Years of Solitude" is
> particularly bad, this type of statement goes at the Expression level.
> But will they know that without help? I'm not sure how to handle this
> sort of thing from a user interface perspective.

You are right that this should be at the Expression level, but I  
despair of ever having WEMI understood by the general Internet  
population when even catalogers are struggling with it. You are also  
right that if there is going to be any understanding of it, the UI is  
key. I'm fairly comfortable with the choice made by Open Library to  
develop Work entities, and to combine Expressions and Manifestations  
as a single entity. For texts, the Expression has few properties, the  
primary one being language/translation. It is possible that if we use  
the role properties (such that author, translator, conductor, director  
would be roles that link an Agent to a resource), we can determine  
that certain roles are Work level and others are Expression level, and  
we can possibly infer expressions. But how to get that note pointing  
to the Expression and not the Manifestation or Work... I'm not at all  
sure about that.

kc

>
> Cheers,
> -w
>
>
> --
> William Waites           <william.waites at okfn.org>
> Mob: +44 789 798 9965    Open Knowledge Foundation
> Fax: +44 131 464 4948                Edinburgh, UK
>
> _______________________________________________
> open-bibliography mailing list
> open-bibliography at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/open-bibliography
>

-- 
Karen Coyle
kcoyle at kcoyle.net http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet