SUO: *Date 13 Feb 2002
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
SUO WG Members,
I am returning, most likely a bit intermittently, to a more general
consideration of the conformance question. Presently, I am working
with the following adaptation of the SUMO group's proposal:
| Conformance
|
| Implementations of S (= SUO) are "ontologies" or "information models".
| A conforming implementation is an ontology or information model Q
| such that all three of the following rules are satisfied:
|
| 1. For every term t occurring in both Q and S,
| the axioms for t in Q must be exactly those
| axioms for t in S whose component terms t_j
| all occur in L(Q).
|
| 2. Every term t in Q is such that:
|
| a. t appears in S,
|
| OR
|
| b. t has axioms in Q,
|
| AND
|
| the axioms that t has in Q are sentences of L(S),
|
| OR
|
| when the axioms that t has in Q are translated from L(Q) to L(S)
| they become sentences of L(S) that contain only the most specific
| appropriate terms t_j that are axiomatized in S,
|
| OR
|
| c. t is axiomatized in (or using) terms that have property 2b.
|
| 3. Q is internally consistent and it is consistent with S,
| that is, a contradiction cannot be derived by means of
| first-order logic from the set of statements belonging
| to either Q or S.
|
| Adapted from "IEEE Standard Upper Ontology (Draft Proposal)",
| Ian Niles and Adam Pease, Teknowledge Corp., 29 January 2002,
| http://ontology.teknowledge.com:8080/rsigma/FormalSUOdraft.rtf
I am also sampling the textual material that
I found by following the following pointers:
http://ontology.teknowledge.com/
http://ontology.teknowledge.com/cgi-bin/cvsweb.cgi/SUO/Merge.txt
http://ontology.teknowledge.com/cgi-bin/cvsweb.cgi/SUO/Merge.txt?rev=1.26&content-type=text/x-cvsweb-markup
The references that I presently have for KIF are here:
http://suo.ieee.org/suo-kif.html
http://logic.stanford.edu/kif/dpans.html
I would like to start out thinking about this material in purely
syntactic and proof-theoretic terms. I may have forgotten how
to do this "purely" enough, so give me some time to remember.
From that perspective, I need to begin with the following questions:
1. With regard to Clause 1, stating that implementations of the prospective
standard S (= SUO) are "ontologies" or "information models", what does
the actual text of an "ontology" or an "information model" look like,
and, less purely speaking, how is it intended to be taken?
I have been assuming up till now that an "ontology" in this sense
is supposed to be "first order theory, more or less" (FOTMOL?),
presented in the form of a "finite axiom set". Is this right?
I still do not get, probably from a lack of sufficient reading,
what the definitive text for an "information model" looks like.
Is it a theory, presented by axioms? Or what?
2. With regard to Clause 3, I cannot begin to examine this in a serious way
without knowing the precise form of the inference rules or schemata that
are intended to be used. I could find no substantive information about
the intended rules of inference in any of the above listed documents.
That seems like enough for now.
Jon Awbrey
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤