Title: RE: [regrep] Minutes of today's meeting
David,
I think I can help from the UBL side of things (you may already know some or all of this):
<Snip>
I would therefore suggest that UBL can take a similar contextual
approach and show how the components are derived using
a layered business use approach.
</Snip>
UBL has a "Context Methodology" subcommittee (http://www.oasis-open.org/committees/ubl/cmsc/) that is covering this area. Their purpose (from their page) is "to develop a methodology and tools for applying context to the core library of generic business information entities (BIEs) in order to produce contextualized BIEs, and develop initial machine-readable descriptions of context rules, in the service of helping the Library Content SC do its work."
There are also 2 other subcommittees dedicated to context:
- Context Drivers
- Context Assembly
The Context Drivers SC is on the verge of releasing a draft (their first, I believe).
I am a member of the Library Content SC (http://oasis-open.org/committees/ubl/lcsc/), chaired by Tim McGrath. The CIQ Team reviewed our first release (Order) back in April - here is an e-mail excerpt from Ram Kumar (4/8/02):
"The CIQ team also believes that there are overlaps between what CIQ is
doing (eg. international name and address definition) and what the UBL
team has come up with to represent customer centric data. This is where
we see how CIQ definitions for customer information can be re-used by the
UBL team."
Hope this helps,
Joe
> **************************************************************************
> Joseph M. Chiusano
> Logistics Management Institute
> 2000 Corporate Ridge
> McLean, VA 22102
> Email:
> Tel: 571.633.7722
> **************************************************************************
>
-----Original Message-----
From: David RR Webber - XMLGlobal [mailto:]
Sent: Friday, August 09, 2002 1:58 PM
To: CHIUSANO, Joseph
Cc: 'Nikola'; ebXML Regrep (E-mail)
Subject: RE: [regrep] Minutes of today's meeting
Joe,
As ever - insightful and precise! YES!!
The approach CIQ is taking is to use ebXML assembly to detail the
contextual use overlayed on top of the finest grain detail.
Therefore you can use the syntax at the level of detail you need.
This is vastly necessary in address - where you have approx' 207
countries today with postal services and about 5 major address
formats per country - give or take - depending on the context of
the address. An example is a loading bay at a super-store, as
opposed to a POBox as opposed to a cubical in a skyscraper office.
The key thing here though is the underlying dictionary and semantics.
And the standard should nit attempt to define the thousands of these
permutation - rather provide the buildinng blocks and the mechanisms
to express and share these 100% accurately with trading partners,
focused around ebXML standards.
This approach is allowing a USPS address to be defined in terms
of the CIQ - and then good things like interoperability can flow
internationally in terms of structuring and using addresses
accurately and consistently.
I would therefore suggest that UBL can take a similar contextual
approach and show how the components are derived using
a layered business use approach.
This is getting us somewhat off topic - except to say that
ultimately the registry needs to support these use patterns
so would be good to base the registry content model on the
same lines.
My suggestion would be to review this in Geneva next month,
as the assembly work is being formalized and aligned with
the other groups - and then post-Geneva we will be able
to do a clear impacts and issues review.
Thanks, DW.
============================================================
Message text written by "CHIUSANO, Joseph"
>
David,
Thanks for your thoughts - I always appreciate and value your insight.
I wonder if the requirements of the 2 representations of "Organization" are
vastly different, due to their purposes. For a vocabulary such as UBL,
there may be a need to identify an Organization to a finer level of detail
then a registry. A registry may require only very high-level information
(as our RIM does, with Address, Parent, Primary Contact, and Phone #'s),
while in a transactional environment it may be required to exchange more
detailed Organizational information between parties (such as the countries
in whicih the organization does business, etc.).
Because of its purpose, the CIQ information is (as you know) very
fine-grained. I would not imagine that in a registry, one would want/need
to include such fine-grained name and address information (it might be
overkill).
Regards,
Joe<
