Regarding this recent thread, I thought it might be useful to toss in
my 2 cents with respect to my recent involvement in the Monticello
project. In essence, DC is the semantic framework for multiple
distributed databases serving differing types of resources.
Nutshell version: I highly suggest separating SUBJECT and DESCRIPTION.
<cents amount=2>
Ok... As I see it, for descriptive purposes, one element
(e.g. Description) with appropriate scheme and type qualifiers may
indeed be enough, for example:
<Description type=subject scheme=lcsh>controlled LCSH subject
headings of the resource</>
and
<Description>free flow of text describing the resource</>
and (for the image-friendly)...
<Description type=generated scheme=x-histogram>red:66%; green:44%;
blue:23%</>
adequatly reflect the differences between controlled and uncontrolled,
differing types of description...
>From a discovery model, however, if I want to search or browse these
values, it seems that I very much want to separate these out. I
absolutely agree with Ray's recent post... Users (generally) don't
want to browse on free flow of text, but I very much may want to for
controlled subject headings... From a discovery point of view,
separating these seems to be very important.
In light of this, (and indeed careful of element-name-feature-creep)
I too highly suggest separating SUBJECT and DESCRIPTION.
</cents>
eric j. miller <URL:http://purl.oclc.org/net/eric>
[log in to unmask] Office of Research, OCLC, Inc.
[log in to unmask] Dept. of Geography, The Ohio State University
|