Many thanks Andy your and Rolands replies really helped
in clarifying some things
>
> My personal preference is for approach b). Note this isn't
> an official DC
> position, but I think it is very much in line with the
I took this to be the most likely direction which brought
me to the problem of not having enough space to encode in DC value of
knowledge organization systems that consists of label/description or
even some kind of taxonomical structure. But this is what Roland
solves neatly with his dcq xml/rdf encoding
I'll browse through DC-Agents and DC-Architecture archives to see
whether
these issues are covered there in more length.
Thanks for dc-xml-guidelines URLs, will read this.
> between metadata records, I guess that the DC-Architecture
> list is the
> better place - one will also want to link to 'authority' files for
> dc:subject, dc:coverage, etc.
I also hope to find out how do they link dc:subject to dc:coverage :-)
as coverage is only facet of subject and they may use the same
encoding scheme. E.g. is one going to repeat 'France' in both places
if
the subject of document is 'economy of France' and coverage of subject
is 'France'. Speaking of redundancy ...
anyway many thanks
cheerio
aida
|