Part of an anonymous discussion: 12/17/2003 8:02 AM
One
scientist says: The justification of the Knowledge Technology uses our sense of what
it might mean for human cognitive acuity to be brought more directly in contact
with representations of actual structure in text of other data
The other scientist says: If
coming more directly in contact with the structure results in reducing ones ambiguity
with regard to the knowledge conveyed by the artifact, then yes, my sense is
that this should improve knowledge discovery and transfer.
However, I still think there is much we don't know about how differences occur between the implicit knowledge conveyed by structure and the explicit knowledge intended to be conveyed by the text or data.
We know such differences to occur -- sometimes intentionally, sometimes by mistake -- so the question becomes why. Which aspects to this question should be viewed with greater importance, the knowledge gained through interpretation of the explicit or the implicit. Along those lines, I do believe that improvements in ontological analysis can help answer questions of this type.
(The host says:
The assumption here is that the enumerated aspects of this question can be
ranked in a list ranked by some measure of importance.
This issue of ranking has
been addressed by Prueitt’s famous question “Which of my three daughters do I
love the most?”
Certainly we must all
agree that there are times when a reduction to such ordered priorities is
simply not appropriate. In these cases,
we often have clear natural ambiguities, and the ranking – any ranking – is an
illusion and results in false sense making.
The ordering is through
the use of an abstraction – called counting. )
The first scientists says: Yes, the work on categoricalAbstraction
and eventChemistry depends on making
this “mistake” and to treat things as ordinal when in fact they are not
ordinal. The work on formative and
differential ontology uses both
an implicit mathematical model (using Hilbert space transforms that cannot be
fully represented on a computer) and an explicit model using ORBs. How the models come into existence and how
are they reified (made human like) through cognitive inspection are addressed
in the first two aspects of the nine aspect Actionable
Intelligence Process Model
(2002).
The method of
ambiguation/disambiguation is
what we offered to the NdCore technology, but they failed to appreciate this to
the point where they would pay the knowledge scientists to integrate a
systematic methodology to reconcile the terminological differences that various
communities of practice will hold dear.
Schemalogic also addresses this same issue. The issue is a under constraint on the interpreted meaning of
words and phrase, even gestures. This
is a feature of natural language which when studied in the abstract cannot be
reduced to the type of logical (closed form) formalism that one finds in
machine ontology like RDF (Resource Description Framework) based OWL (Ontology
Reference Language) or the Cycorp ontologies.
However, the method of
ambiguation/disambiguation using Subject Matter
Indicator neighborhoods does
directly address this matter. It
addresses it in two ways.
First, the Ontology Referential Base (ORB) is developed using a
computer implementation of the notational paper (completed in November 2003). The results of this implementation is then used to produce a
“datawh.txt” file for the SLIP browsers, which then creates the ORB and allows
a trade secret to identify the center level of what is a multi-level
fractal. The patterns in this layer are
then stood up as Fixed Upper
Taxonomy.
Finally, human cognitive acuity and tacit knowledge is used in the
development of a refinement of the Subject Matter
Indicator neighborhoods.
The result are easy to use,
completely independent ontological structures that are stored and viewed as
very simple XML file, but which can be rendered as OWL or Topic Maps. A provisional patent was filed on December 1st
and then published at:
http://www.ontologystream.com/beads/nationalDebate/sixteen.htm