On Mon, 16 Aug 1999, John A. Kunze wrote:
> Dear Dublin Core community,
>
> Would any of you have any numbers regarding DC usage in the library,
> museum, archive and academic worlds? I'm looking for any rough measures
> of saturation, e.g., "2% of the union of all html pages from library
> websites globally contain DC metadata tags".
We do have some statistics for the *.dk and *.se domains. But the figures
are not really in the form you want it. I reckon that Lund university has
about 200,000 WWW pages. I don't think more than 2,000 of those are DC
labelled, out of which 1157 have been labelled within a project I'm
working on. So, let us say that between 1-2% has proper DC metatags in
them.
The bulk of these pages has been meta labelled during the last 12 months
or so. To give you a rough idea on the situation in Sweden in general I
can give you a crude idea on metadata usage through thes table included
below (numbers in left column - sample taken 17 June 1999 and it is biased
since it excludes Swedish pages in some popular domains, notably *.com and
*.nu). If we include keywords, author and description in what we regard as
useful metadata, then the situation changes dramatically. But please note
that the occurence of DC metadata is now in the _same order of magnitude_
as the occurence these, socalled Alta Vista tags.
I would like to stress one further point. I have never expected that the
fraction of properly meta labelled pages would increase beyond 10% or so.
Using DC metatags is a kind of cataloging. Librarians estimate that, in
Sweden with a population of about 8 million, they do first time cataloging
of about 16,000 objects a year. Swedish WWW authors metalabelled about
8,000 WWW pages last year. The metatagging effort is still increasing, but
even with its current volume it is already today the second or third
largest single cataloging effort in the country! I think it this in a
fairly typical description of the Nordic Countries in general.
Yours
Sigfrid
485954 Records with any metatags whatsoever
361952 generator
82412 keywords
66705 author
54814 description
39641 formatter
24373 dc.publisher
24328 dc.title
21152 dc.publisher.address
12142 template
10838 dc.language scheme=ISO639-1
10319 dc.type
10192 dc.description
9934 safari.targetgroup
9186 publisher
7774 dc.format
7599 dc.relation
7466 dc.relation.ispartof scheme=URL
7466 dc.relation.ispartof
7438 dc.creator.corporatename
7265 microsoft
6581 dc.relation.ispartof scheme=URN
5995 dc.creator.personalname
5911 distribution
5561 dc.subject
5303 resource-type
5256 ekdoctech
5255 ekbu
5253 ekdocowner
5207 ekreviewdate
5175 ekdoctype
4722 dc.date.x-metadatalastmodified scheme=ISO8601
> I know it's very hard to measure and interpret such figures, but any help
> in this direction (e.g., pointers to studies) would be appreciated.
>
> -John
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|