SUO: Re: Building the Hierarchy
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
Pat,
Still working my way through Robert's and John's notes,
but since it's Friday, and since I have some things
off the top of my head to say on this issue, I will
skip on ahead. Strange to say, I have actually
spent (wasted) some time thinking about this.
The ":" is habitually read in
math and comp sci as 1 of 2
related things:
1. "of type", as in f : X -> B
2. "such that", as in f : f(x) = x^2.
The "::" is used in proportions, as A:B :: C:D,
but that is probably not too much of a problem.
A bigger issue is this: When possible, it helps to use
symmetric signs for symmetric relations, and asymmetric
signs for asymmetric relations. The fact that ":" already
violates this actually causes some expressive problems in
applications of type theory. To work around this, I use
"<:" for "is of the type" and ":>" for "is the type of",
in such a way that x <: T iff T :> x.
I am guessing that you want to put the agent name or community name
first because you are thinking of directory trees or namespace trees.
If it were possible to enforce mutual exclusion or non-overlapping
contexts among the variant usages of the same stem (rather, "leaf")
then this would be a natural stratagem. If you think about it,
I think that you will see that this is only possible when the
diverse interpreters do not really talk to each other at all.
That is why I think of the "term to interpreter" relation as best indicated
by an "index" or "subscript", since it reminds us what is commonly called
the "indexical" properties of signs, and the diverse belief systems to
which various communities of interpretation "subscribe".
Jon Awbrey
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
Patrick Cassidy wrote:
>
> One suggestion about naming conventions:
>
> > The first point to note is that every module in
> > the hierarchy should have a unique identifier.
> > Therefore, all names could be distinguished by
> > a concatenated string such as "moduleID.localId".
>
> I would like to use the name of the Ontology from which
> each concept comes as a prefix with a double colon, thus:
>
> SUMO::Object
>
> > . . . to provide each concept a unique identifier.
>
> Is this naming convention OK?
>
> Pat
>
> =============================================
> Patrick Cassidy
>
> MICRA, Inc. || (908) 561-3416
> 735 Belvidere Ave. || (908) 668-5252 (if no answer)
> Plainfield, NJ 07062-2054 || (908) 668-5904 (fax)
>
> internet: cassidy@micra.com
> =============================================
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o