Re: [xml-dev] Musings on the fundamental nature of data

["Timothy W. Cook" <timothywayne.cook@gmail.com>]

On Mon, Apr 29, 2013 at 2:26 PM, Costello, Roger L. <costello@mitre.org> wrote:

>
> Assertion: The most fundamental issue with data exchange is:
>
>     What worldview do you want replicated?

I fully support this assertion.

In some (most?) domains though these are not static views.

Specifically in the domain of healthcare (ignoring the fact that
domain experts prefer you to just suck the knowledge from their
brain).  When dealing with the healthcare record of a single person
over their lifetime, it is vitally important to capture 'data' in the
context that it was created combining the ontological, temporal and
spatial contexts along with the data into an artifact that can be
called information. Static data models do not work in this environment
because not only does time pass, the patient changes and is in various
geographical contexts. The science changes as well.  You cannot
migrate the data into new ontological contexts and still have it be
valid as it was captured.
This has led to the multi-level modelling approach using a generic
reference model and domain models that can vary over time.

I'm sure this can be helpful to other domains but man made domains do
not seem to be as complex as evolutionary domains.  Evolution has had
millions of years to work on the models and they are designed to be as
complex as necessary whereas man made models are designed to be as
simple as possible.

--Tim