Re: vocabs, metadata set, datasets

In unimarc authorities, there is a notion of "primary entity" to account for "the thing". I think the same could be believed for bib records.

Jeff

Thomas Baker <tbaker@tbaker.de> wrote:

On Tue, Jan 18, 2011 at 02:33:08PM +0000, Gordon Dunsire wrote:
> A definition of "record" from the MADS/RDF Vocabulary Description document
> (http://www.loc.gov/standards/mads/rdf/):
>  
> "Used throughout the current document to describe the compilation of facts about
> something or, especially in the case of the present document and topic, the
> collection of triple statements about an authoritative or variant term (such as
> a subject or name heading)."
>  
> Which generalises to: "The compilation of facts about something, or the
> collection of triple statements about something."

Put this way, it sounds to me like what the DCMI Abstract Model
calls a Description (see below).

> See also the Background section of the BibData use case cluster
> (http://www.w3.org/2005/Incubator/lld/wiki/Cluster_BibData): "... bibliographic
> record, a set of data elements describing the content and characteristics of an
> information object manufactured for human consumption.". Which generalises to "A
> set of data elements describing something." - but are we developing a definition
> of "record" within the bibliographic context, or just generally?

I see the argument but still see a problem with the notion of
a record being "about" something.  Looking at a bibliographic
record through traditional bibliographic glasses, one does see
a set of data elements "about" a book.  Looking at that same
record through RDF glasses, however, one may see statements
"about" several different things -- the book, its author,
the publisher, etc.

The DCMI Abstract Model tried to express this point by
distinguishing between a Description (a set of statements
about just one resource -- e.g., the book OR the author OR
the publisher) and a Description Set -- a set of related
descriptions bundled together for practical purposes in a
single metadata "record".

Yes, we need to use a terminology related as recognizably as
possible to things already familiar to the intended reader,
but we also need to point out to readers where perspectives
differ.  On this point, I suggest we find a way to say
that "records" may provide descriptive statements "about"
more than one something.

Tom

-- 
Tom Baker <tbaker@tbaker.de>

Received on Wednesday, 19 January 2011 20:03:13 UTC