SUO: Re: Evaluation Framework for Content Standards
----- Original Message -----
From: "John F. Sowa" <sowa@bestweb.net>
To: "Erik Larson" <elarson_78746@yahoo.com>
Cc: <standard-upper-ontology@ieee.org>; "Nicola Guarino"
<Nicola.Guarino@ladseb.pd.cnr.it>; <cg@cs.uah.edu>
Sent: Monday, June 16, 2003 10:39 AM
Subject: SUO: Evaluation Framework for Content Standards
[snip]
> I believe that we will have to make many decisions, primarily
> on the choice of distinctions that can best accommodate every
> conceptualization of the world that anyone might require for
> any purpose whatever. Those choices are empirical issues that
> depend on examining the evidence. A methodology, such as formal
> concept analysis (FCA), can help by showing the consequences
> of any decision. And FCA is one of the formalisms that are
> packaged in the IF framework.
Here is a possible methodology. With n binary either-or distinctions there
would be N = 2^n ways to partition the "sorts"of things. Distinctions
correspond to conceptual scales in FCA. Either-or distinctions correspond to
dichotomic scales. But there are many other kinds of conceptual scales,
including nominal scales, ordinal scales, interordinal scales, etc.
Combining or resolving in terms of distinctions corresponds to conceptual
scaling in FCA. References to conceptual scaling literature were given in
http://grouper.ieee.org/groups/suo/email/msg09607.html
Analysis, resolution or working top-down:
* Take any theory T with underlying language L and any partition of the
entity types (sorts) of L into say N blocks.
* Partition the relation types of L according to whether all the entity
types in their arity (argument types) are in a particular entity type
partition block (EB).
*There will be N+1 relation type blocks (RB), one for each EB plus an
addition block for those relations whose arity spans more than one EB.
* Partition the axioms similarly.
Ignoring the extra spanning block, this creates N sublanguages of L and N
(more generic) theories above T within the lattice of theories over L.
* To get the original theory T, take the meet (union) of these N theories
plus the specialized theory containing the spanning relationality and
axioms.
Synthesis, combination or working bottom-up.
Generalize the steps implicit in the John Sowa's library of modules diagram
http://www.jfsowa.com/figs/suohier2.gif,
or explicit in the language module processing steps
http://suo.ieee.org/IFF/language-module-processing.html
Robert E. Kent
rekent@ontologos.org