SUO: Re: Now What?
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
Graham Horn wrote:
>
> Jon et al,
>
> I agree with your perspective that "an ontology is like
> a single data structure". However, I believe you are
> underestimating the power of the modern entity and
> relationship(/process) analytical processes available.
> Some of this is covered at <http://www.idef.com>
> especially IDEF0 (process) and IDEF1X (data),
> now also combined in "Object Role Modelling"
> championed by Terry Halpin; but also now
> IDEF5 for ontologies.
>
> I suggest what you are calling "a single data structure"
> is only a portion of one.
Size does not matter.
It is their motility,
and the lack thereof.
> For example my latest info is that one portion
> of the US Department of Defense has a data model
> that cannot be viewed in a single diagram, and the
> multiple perspectives take several walls to display in
> a readable form. They have over 10,000 data definitions.
Good example. And do you suppose that "Chinese Embassy in Belgrade"
is borne somehow somewhere in this Mother of the Ten Thousand Names"?
What does it matter if you can't fold the map?
What good is it if nobody asks for directions?
> This perspectives aspect touches on what I have been
> referring to as dimensionality so many times in this project.
>
> I suggest you become familiar with the complexity accommodated by
> proper data (and process) modelling, etc. Then you will answer
> your concern about its ability to model generic situations.
I suggest you go teach your grandmother how to suck eggs --
sorry, I've been taking lessons in plain speaking lately --
it's kind of a work in progress.
> Sure, it will be an incremental process,
> just like a young child learning to speak.
> But it will grow.
So does fungus.
Look I've kept up in AI long enough to see
each new generation go for that "toss the
nuts'n'bolts in a bucket and shake it
until you get a robot" routine.
It ain't evolution --
It ain't development --
It ain't even research --
It's just wishful thinking.
Cheers,
Jon
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤