"Dynamic schemas" is the way I am capturing this thread...representing
different perspectives based on established patterns.
Here is an instance of a perspective (i.e.
authority) 'module' I came up with-- for lack of a better word at this
stage.
These can describe each XML Schema. RDF is used to annotate XML
documents, but has it also been used for annotating XML Schemas
themselves? I still have some past threads and reference to
study regarding RDF related to HumanML, but this is a question that now
pops up for me.
The 'bias' factor has to made explicit at the onset.
<perspective type="??"
date="09.02.2001" extrapolation_type="direct inquiry" />
<individual>Bill
Jacobs</individual>
<individual>Representative John Billy</individual>
<individual>Senator Joe Whiter</individual>
<group>US
State Dept</group>
<group>Canadian
Gov't</group>
</perspective>
DTD (for simplicity sake, for now)...
<!DOCTYPE [
<!ELEMENT perspective (individual*, group*)>
<!ELEMENT individual #PCDATA>
<!ELEMENT group #PCDATA>
<!ATTLIST perspective type CDATA #REQUIRED>
<!ATTLIST perspective date CDATA #REQUIRED>
<!--Choices below include assumption, direct_inquiry, ModelX,
ModelY, algorithmX, patternZ, or anything else-->
<!--It may describe how the perspective was
derived-->
<!ATTLIST perspective extrapolation_type CDATA #REQUIRED
"assumption"
]>
(of course, digitally signed and verified through a mechanism--possibly
such as what Sean was developing)
I don't think in a practical sense we will need to deal with so
much complexity as Paul's research was directed towards--at least
to get the initial perspectives flushed out (although time
will tell).
I don't think we will need to get too abstract either.
In other words, I don't think we need to establish abstracted pattern
matching models to describe perspectives, or utilize mathematically tranform
perspectives (Len: when you use the word stylistic modifications, I am
assuming you mean transforming through XSLT correct?)
It is much better to let the humans themselves define them directly, as
Rex has been emphasized previously, through individuals
themselves. As humans, we polarize towards what is concrete
anyway, for better or for worse. If authority is clear and
unequivocal, we start to share a common perspective. Patterns start to
merge and come together, and the complexity relating to differences may
not seem so complex.
Too early to say though how it'll evolve.
-----------
Ranjeeth Kumar Thunga
----- Original Message -----
Sent: Friday, September 14, 2001 12:40 PM
Subject: RE: HM.Requirement: authority (Re: Case in Point-cultural
Module:IslamicFundamentalism?)
>
> And to understand viewpoint, one must
understand it changes
> with respect to distance from that which is
viewed. Viewpoint
> has dimensions and the schema attempts to
capture these such
> that the principles of focus can be brought to
bear. A
> ball of twine is a point from a distance, a sphere
from a
> distance, a cylinder from a distance and viewed on end, a
> filled circle and a point. The trick is to understand
> it as a pattern. This is what Prueitt is pointing
out.
>
> The pattern directs cells of process. He
calls them
> process compartments and while referred to using other
> terms in earlier works, that is good enough. The notion
> of orchestration is useful because it takes in the
> idea of
addressable process types on a timeline with
> some variation
possible given stylistic conventions.
>
> Schemas are
patterns, can be dynamically adjusting
> by stylistic convention, and
are directable using
> well-configured processes. One of the
processes is
> identification of source and type.
>
>
Len Bullard
> Intergraph Public Safety
> clbullar@ingr.com
> http://www.mp3.com/LenBullard
>
> Ekam sat.h, Vipraah bahudhaa vadanti.
> Daamyata. Datta.
Dayadhvam.h
>
>
> -----Original Message-----
>
From: Kurt Cagle [mailto:kurt@kurtcagle.net]
> Sent: Friday, September
14, 2001 11:30 AM
> To: Ranjeeth Kumar Thunga
> Cc: slbain@netobjectives.com; humanmarkup-comment@lists.oasis-open.org
>
Subject: Re: HM.Requirement: authority (Re: Case in Point-cultural
>
Module:IslamicFundamentalism?)
>
>
> This is an
incredibly good idea. Schemas by definition impose a viewpoint;
> as I
point out in most of my XML classes:
>
>
**********************************************
> You cannot understand
the fundamentals of schemas without appreciating the
> fact that all
schemas are intrinsically political, not technical. If you
> have
three departments that each have a need for a personalization
schema,
> then each department will have its own priorities and
interests for that
> schema. In that regard a schema is in effect a
social contract, an agreement
> between all parties about the domain
of the schema, the terminology used,
> the relationships (and relative
priority) of elements within the schema.
>
**********************************************
> I suspect that within
a decade, schemas will likely end up becoming an
> integral part of
all civil legal processes - are you trying to build a
> building? Then
you agree to use architectural schema
> http://www.archstandards.gov/schemata/bld1254a6
and permit process schema
> http://www.archstandards.gov/schemata/prmt2399ds
. These become part of the
> legal records, and are kept as part of an
online repository.
>
> As I see it, a significant part of what
the HumanML group itself is trying
> to do is to create a set of
schemas that attempt with some modicum of
> fidelity to model aspects
of human behavior and interaction. This is of
> course not fully
possible, precisely because in the creation of such schemas
> we do
create a bias, but if we can recognize that from the outset and
>
attempt to mitigate the bias (or introduce some mechanism that makes
it
> possible to change the bias while still maintaining fidelity to
the
> standard) then I think we can go a long way toward building a
more universal
> schema.
>
> -- Kurt
>