On 9/7/07, Riley, Jenn <jenlrile_at_indiana.edu> wrote:
> Can't we
> collaboratively design database structures that conform to FRBR
> principles but also meet system needs of normalization, cataloger needs
> for streamlined data entry, etc., etc., etc.?
Sure we can. But why not aim even higher? Why not just adopt Topic
Maps now, get FRBR for free (reduce it to an ontology task instead of
implementation task), data normalization not needed (not as a modeling
exercise), persistent identifiers, merging across domains / records,
true semantic modeling, all wrapped in lovely XML if you prefer. Or,
I'll be happy with RDF as well. Any rich exchange format in front,
where legacy is controlled separately, and where the user interface
isn't tied to technology at all.
> Can't we at least consider
> options beyond a 1:1 match between the data structure and the conceptual
> model, if even to, in the end, determine it's better to stick with the
> 1:1 match?
There's always pragmatic decisions behind conceptual, user and data
models all looking alike. Most of the time it has to do with
resources, but often about technology. The IT world has spent the last
30 years trying to pry these models apart, and there's tons of
possibilities, but none as long as our vendors aren't part of that
solution. I'm wondering where they are in all of this. I see a couple
of guys from Talis. Anyone else out there lurking, and want to give
customer service? :)
Alex
--
---------------------------------------------------------------------------
Project Wrangler, SOA, Information Alchemist, UX, RESTafarian, Topic Maps
------------------------------------------ http://shelter.nu/blog/ --------
Received on Thu Sep 06 2007 - 20:04:30 EDT