Re: Tim Berners-Lee on the Semantic Web

From: Karen Coyle <lists_at_nyob>
Date: Tue, 20 Oct 2009 08:12:19 -0700
To: NGC4LIB_at_LISTSERV.ND.EDU
Houghton,Andrew wrote:
>
> Personally, I believe that FRBR and RDA will be more useful to the 
> Semantic Web since they are both based on a domain model which can
> be expressed as an RDFS/OWL ontology.  People can argue whether the 
> domain model is correct, but having a domain model is important to
> understanding the data and its relationships within the domain model
> and outside the domain model, e.g., linked data.
>
>   

I think we've gotten to the point in these discussions where we have to 
define our terms. In the same way that "MARC" means both a structure and 
data elements and a content standard, the terms FRBR and RDA are now 
taking on multiple meanings.

As produced by the JSC, RDA is a huge text with rules for how to decide 
the elements of a catalog entry to describe an resource and its access 
points (except for subjects). That huge text has been turned into an 
interlinked XML document that will be offered through an online system, 
RDAonline. The observed data elements from the RDA text have been 
registered as RDF-compatible properties in the Metadata Registry. 
(http://metadataregistry.org) The controlled vocabularies are also 
registered there, under the "Vocabularies" tab. This means that all of 
the data elements and the various lists (Colour of Moving Image, Groove 
Width) are there and exportable in a machine-readable form. Note, 
however, that there isn't at this time an actual record format for RDA, 
although if one develops I sure hope it uses the RDF properties and not 
some older technology.

There is no particular reason why one couldn't create a set of 
RDF-compatible properties for AACR2, or even for the MARC fields and 
subfields -- although RDA conceptually seems closer to the semantic web 
goals than AACR2 or MARC. I think things turned out the way they did 
mainly because RDA, as conceived by JSC, is promoted as a "new" set of 
rules, so there was a nice void into which one could try to introduce 
some additional changes in thinking. RDA does have a lot of legacy from 
previous rules and appears to have conceded in some areas to concerns 
that it will have to fit into the basic MARC structure. This is why, 
AFAIK, we have things like "publication statement" defined as an element 
in RDA, even though you could imagine using place or date of publication 
as simple data elements under some circumstances. Because of these 
legacy concepts, RDA as defined in the Metadata Registry is a bit kludgy 
-- a "pure" semantic web representation of bibliographic metadata would 
probably be cleaner.

FRBR is in a similar situation, although lagging a bit behind RDA. IFLA 
has committed to creating an RDF-compatible version of FRBR and FRAD 
(and any other FR's that come along, presumably) and adding these to the 
Metadata Registry. This would then allow one to apply FRBR structure to 
RDA properties, and create bibliographic information that adheres to 
FRBR and is compatible with semantic web data structures.

We need some handy and clear terms to refer to RDA and FRBR in their 
machine-actionable, RDF-ized forms as opposed to the original text 
documents where the concepts were explained. The RDA elements are 
registered under the domain "RDVocab", which could be one way of 
distinguishing those elements from the JSC RDA text.

kc

-- 
-----------------------------------
Karen Coyle / Digital Library Consultant
kcoyle@kcoyle.net http://www.kcoyle.net
ph.: 510-540-7596   skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
Received on Tue Oct 20 2009 - 11:15:49 EDT