Back ...
... ... ... ... ... ... ... ... ... ... Send comments into the ontology
stream. ... ... ... ... ... ... ... ... ... ... ... Forward
(:..:) ß different threads à (:….:)
Bead 1
January
1st, 2001
OntologyStream Fable Indexing Project
Bead 1
January 6th, 2001,
The OntologyStream Fable Indexing Project is intrinsically very exciting, as it is oriented towards
representing the semantic linkages within and in between 312 Aesop fables.
The notion of semantic linkage is to be determined during the
Project and will reflect a Peircean (Charles Sanders Peirce) view of semantic
composition. This means that human
judgment is needed initially to build a theory of semantic type and valence for
the collection.
Once humans have worked out a theory about how semantic linkage
reveals itself in the fable collection, then the OntologyStream will allow
various technology companies and academic groups to test technology on
Indexing, Routing and Retrieval (IRR) related to navigation in the
collection.
The “ground truth”, for the tests, will be a hand constructed
semantic graph consisting of a collection of syntagmatic units having the form <
a, r, b > where a and b are passages in the fables and r is a relational
variable. The notion of a syntagmatic
unit comes form the work of D. Pospelov in Russia and is consistent with an
experimental ontology streaming system using Mill’s logic
and Peircean logic.
We are looking for an academic group or literary individual who
would go through a specific process in the development of a hypertext document
containing 312 fables. In the hypertext
document the links are to be categorized so that paths from one fable element
to another fable element have quirk properties (color and spin and other
differentiators, as in quantum mechanics.)
Quirks should represent the semantic valence between the fable
elements.
Perhaps we can call this “quantum linguistics”.
Our methodology for finding all semantic quirks in fables is to
descriptively enumerate:
1) first the fable
elements, a, b, c, d, ….
and second the
2) the semantic linkage, {
r }, between any two pair of these
elements
A fable element is any portion of text that is in essence a
passage. Passages need not be
contiguous and the set of all passages need not partition the text into
disjoint passages. The category theory
assumes that passages naturally over-lap.
The linkage can be identified as being present without the need
to classify. The development of a
theory of semantic linkage should be at least as complex as the taxonomy of
subatomic particles. The method of
descriptive enumeration, however, poses the prospects that a degree of
arbitrariness is not only allowed but a necessary part of our process of
constructing a simple semantic linkage taxonomy to serve as “ground
truth”. Thus we will just do the best
we can, and then use the result to classify the linkages.
Later a full set of possible linkage categories will be
determined by group judgment and falsification logic. These linkages may form a periodic table of oppositional scales
and thus tie into a visualization methodology based on Computer
Cognitive Graphs.
The semantic linkage representation will then serve as a
benchmark for a worldwide review and testing of Indexing, Routing and Retrieval
(IRR) technologies. Several clients
will review the results of the test and acquire technologies based on this
review.
Our present need (January 2001) is for a scholarly group having
an appreciation of the themes within the fable collection. The reward will be the accomplishment of a
navigation framework for moving from one fable to another based on what fables
have been visited recently and a user profile.
It will be a glass
bead game, with various navigation engines. The work, and some of the navigation systems, is to be public
domain and thus research papers will be possible.
Please communicate this bead to anyone who might wish to be
involved in the OntologyStream Fable Indexing Project.
Dr. Paul S. Prueitt
Director BCNGroup
Founder OntologyStream
Knowledge Scientist at Acappella Software