Dear all,
A short report of what we have been doing up here after the DC-4 down
under may be of interest to some of you. I'll not waste bandwith by
going into murky details; in order to get them you can contact people
actually working with them (BTW - Ole Husby will tell more about the
DC->MARC converter when it has been finalised). Another good source
is the project homepage (http://linnea.helsinki.fi/meta/) which has
been updated a couple of times - the latest changes took place in late
June.
1. DC template and user guides.
Shortly after DC-4 Traugott Koch and colleagues updated the Perl
script producing the actual DC record in such a way that HTML syntax
complied with what was agreed on. Nothing dramatic has happened since
then, but there are a couple of things we are contemplating:
Linkage to URN delivery system. While we are currently telling
authors to use URL as (or instead of) Identifier, this will not do in
the long run. We (i.e. the National Library of Finland) are in the
process of working out together with URN-WG a way of coordinating URN
usage on national level. Once things have been worked out on this
level, we can build a system that delivers national bibliography
numbers as URN's to those who need them via the Web. This system can
be easily linked to the DC template.
We are also working on making more Nordic subject headings lists
accessible via the Web.
On a more practical level, we are wondering whether we should
translate resource type names to "native" languages. There are
technical means to accomplish this on the template level, but
searching would be more complicated.
User guides (and evaluation form) were made available in late May. We
think that by October feedback from users and access statistics of
user guide files (each DC element has it's own file) will provide us
with some valuable insights of what are the biggest stumbling blocks
in creating DC-based metadata. Within a month we got 800 hits to the
user guides, and 1300 accesses to the element help pages.
All guides are of course linked to our DC templates, available at
http://www.ub.lu.se/metadata/DC_creator.html
and http://www.ub2.lu.se/metadata/DC_creator_simple.html,
the latter representing "minimalistic" approach to DC creation.
2. DC-aware search service
The Nordic Web Index is well on it's way becoming DC aware. A version
in which DC indexing is integrated into the production system will be
available by October-November this year. At the moment Sweden and
Denmark have already made available separate national metadatabases
for test purposes (see http://nwi.ub2.lu.se/?lang=en). These bases
contain about 30.000 and 50.000 records, but not that many of them
are in Dublin Core or in any other decent metadata format. We intend
to build similar bases to other Nordic countries as well, before the
new NWI version is officially launched.
NWI is also about to become EWI, or European Web Index. England is
the first non-Scandinavian country to build a national web index
using the application (Andy Powell in UKOLN is working on this) and
more European countries will follow later. Finally, Nordic countries are
using NWI to archive harvested web documents (see
http://kulturarw3.kb.se/).
Sigfrid Lundberg from Lund University NetLab Unit
can provide you more information on this task. His email is
[log in to unmask]
3. DC->MARC converter
A beta version of the converter is available at
http://www.bibsys.no/meta/d2m/
The converter can currently produce NORMARC, USMARC and FINMARC
records, either in plain text or ISO2709 (USMARC conversion has not
yet been formally accepted by LC). Other Scandinavian formats will be
added later.
To facilitate the FINMARC convertion, I modified the
LC's crosswalk document by replacing USMARC with our format. A
"nasty" test record was also made in order to choke the converter :-)
(actually it makes a good job in producing FINMARC from this
DC record) at http://linnea.helsinki.fi/meta/dcfinmarc.html
To sum up, it seems that the project can complete all the technical
development tasks listed in the project plan. What we need most now
is active promotion of the project and Dublin Core to potential
DC authors. Luckily there are already projects here, active or in
planning stage, which plan to adopt Dublin Core or DC-like formats,
and Nordic metadata tools. Probably the most interesting of these is
the Danish INDOREG (Internet Document Registration) project, of which
you will here more in the DC-5 here in Helsinki in October.
Thank you for your attention,
Juha Hakala
General dogsbody of the Nordic Metadata project
*****************************************************
Juha Hakala, Helsinki University Library /
Automation Unit of Finnish Research Libraries
fax +358 9 753 9514 internet: [log in to unmask]
*****************************************************
|
