Re: Pizza Example
Date: Thu, 8 Apr 2004 15:24:00 -0500
Message-ID: <c54cdd$upt$1_at_news.netins.net>
"Eric Kaun" <ekaun_at_yahoo.com> wrote in message
news:SRhdc.13831$Uy3.4866_at_newssvr15.news.prodigy.com...
> "Dawn M. Wolthuis" <dwolt_at_tincat-group.com> wrote in message
> news:c53vka$a7$1_at_news.netins.net...
> > "Eric Kaun" <ekaun_at_yahoo.com> wrote in message
> > news:oUcdc.52055$yG6.8836_at_newssvr16.news.prodigy.com...
> > > "Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message
> > > news:eR_cc.63743$eD5.4200565_at_phobos.telenet-ops.be...
> > > > Eric Kaun wrote:
> > > > > "Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message
> > > > > news:YJZcc.63658$fD5.4201522_at_phobos.telenet-ops.be...
> > > > >
> > > > >>Eric Kaun wrote:
> > > > >>
> > > > >>>I'm certainly willing to entertain language notions - I just
> haven't
> > > > >
> > > > > heard
> > > > >
> > > > >>>anything concrete enough to serve as the basis for a data model.
> > Loose
> > > > >>>correspondence to English is, in my opinion, not a good metric.
> > > > >>
> > > > >>Did you ever look at the philosophy behind ORM (Object-Role
> > Modelling)?
> > > > >>Or the philosophy behind predicate logic, for that matter?
> > > > >
> > > > > No, can't say I have in any specific way. Pointers/links?
> > > >
> > > > For ORM see http://www.orm.net and as an introcution perhaps
> > > > http://www.orm.net/pdf/ORMwhitePaper.pdf
> > > >
> > > > For predicate logic you can always read Frege's Begriffsschrift. :-)
> > > > Sorry, only kidding. Can't think of a good reference right now.
> > > >
> > > > The point is that all these *are* in fact based upon a
correspondance
> to
> > > > language. In some sense that was what the great discovery by
Aristotle
> > > > was: the fact that you can sometimes reason on the basis of only the
> > > > *form* of statements. Hence Formal logic.
> > >
> > > Certainly - since computers can't understand meaning in the way that
we
> > can
> > > (which we ourselves don't understand), it follows that form is about
all
> > > there is, and thus the mechanizability of symbolic logic. I'm not
> > > downplaying the role of language, but languages that computers
> understand
> > > are far different than those we understand. So if we're going to use
> human
> > > language as a basis for computing, we have to specify which aspects of
> it,
> > > or subset of it, and furthermore impose rules that might not make much
> > > difference to our comprehension.
> >
> > Obvously, the computer itself need not understand the language in order
> for
> > us to understand language that is output from the computer. So, if we
> don't
> > split up English sentences (for example) unnecessarily when we feed them
> to
> > the computer as "data," then we can retrieve them in a form closer to
the
> > original.
> >
> > So, if "the Pizza has Mozzarella and Parmesan cheese" is stored as:
> >
> > Pizza Mozzarella
> > Parmesan
> >
> > in a single, uh, record, then if the ordering is unwittingly useful, we
> > didn't harm that ordering.
>
> Of course. This assumes that your main reason for storing the data is to
> display it again, in its original form, for humans. If that's all you're
> doing, then many different simple systems will suffice - a Word document,
a
> spreadsheet, etc. But if you're trying to reason about the data, then you
> need to structure it in a way amenable to automated deduction. You are
also
> perfectly free to keep a copy of the original, which would then be
dependent
> upon the key, the whole key, and nothing but the key per normalization
> rules.
I want my cake and eat it too! The PICK structure does what I have described and is "amenable to automated deduction" and it seems to me that there is some value in that, but I'm still poking and prodding to clarify what that might be. The original Pizza example and this example about retaining order (related to another thread) both show up some advantages of the PICK or XML approach vs. the relational approach. I realize they fall far short of proving anything, however. Cheers! --dawn Received on Thu Apr 08 2004 - 22:24:00 CEST