ONT Re: Data Models, Ontologies, Logic

[Jon Awbrey <jawbrey@oakland.edu>]

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

Note 24

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

We are looking at factors affecting the integration of
data bases, full of data points called "tuples", and
knowledge bases, full of sentences called "axioms".

The thing about real data, once you put yourself in the way of it,
is that it just keeps y-cumen in, mounting to a real time onslaught
that is tough to slot in our established knowledge bases, except as
a tacked on accumulation of disjointed disjunctive "axioms" that we
use to express the raw facticity of ongoing experience.  Since many
people reserve the name "axiom" for something slightly more elegant,
succinct, and well-thought-out than that, I will frequently use the
term "constraint" for this particular genus of syntactic "parsicle",
and depending on their use in a given setting I may also describe
their appearance in the role of a "case", a "fact", or a "rule".
More loosely speaking than lucidly thinking, we often describe
the whole knowledge base as a "theory".  There are, of course,
numerous problems with that manner of speaking about theories,
but there will be time to discuss all of that bother later on.

The other thing about real data is its annoying tendency
to puncture our favorite theories of the way things are,
or ought to be.  If you cannot tolerate losing your pet
theories on a recurring basis, you will avoid real data
at all costs.  Look around, I'm sure that you know some
people like that.  I may as well confess that I was one,
once.  Okay, twice.  But I am in my third mind, now.

Jon Awbrey

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o