Chantilly Virginia

Fable Arithmetic Link (in progress)

OntologyStream

Founder: Dr. Paul S. Prueitt

beadmaster@ontologystream.com

Thursday, February 01, 2001

What I am about to say talks a little about several new discoveries.

****

One of these inventions is a “virtual graph database” (my term) technology that will replace the relational database system when the task is complex and involves extensive search space processes.

This technology is the creation of a colleague of mine, Bjorn Gruenwald, and can be demonstrated at any time.

The “virtual graph database” (my term) has its own operating system written by Bjorn in a Forth programming language derivative and runs on any platform. The speed of the complex retrieval processes is five or six magnitudes of order faster than a relational database (working on large data sources).

There is NO precision recall problem – all valid syntactic relationships are retrieved. The true syntactic-relationship-complex (my term) is “grown” (my term) via a patented evolutionary programming process, in such a way that a database that has become corrupted can be cleaned and normalized to recover 100% of all data structures (as defined by the original data model that emerges in a fractal compression type fashion.)

The analogy to fractal compression is correct in that substructure generative elements can be pruned to produce any data model (in fractals any image can be “produced” from substructural fractal generation elements). Thus the “virtual graph database” is a means to “compress” any relational database and then uncompress with vetted biasing. (Really!) The compression, decompression and storage are as a single symbol string (streamed ontology).

A theory of semantic linkage is all that is needed, and this is now discussed (below).

***

Now to my “new” invention.

I will be brief and describe how to justify one aspect of semiotics to the business community. I can now ground a simple description of the complexity of making an interpretation of a sign system (semiotics) onto the work of Stuart Kauffman and the Complex Adaptive Systems (CAS) paradigm developed by the scholars at the Santa Fe Institute.

My description is from a review of work from theoretical immunology (Alan Perelson) that Stuart Kauffman has just made in his new book "Investigations". The basic work is on finite "topological coverings" (my term). I published in this area

Eisenfeld, J. & Prueitt, P.S. (1988.) Systemic Approach to Modeling Immune Response. Proc. Santa Fe Institute on Theoretical Immunology. (A. Perelson, ed.) Addison-Wesley, Reading, Massachusetts.

The aspect has to do with the "objective" existence of a finite number of dimensions. Dmitri Pospelov suggested in "Situational Control" (in Russian, unpublished in English) that there are 117 dimensions to the semantic space - and that this is language independent. He gave me a paper on opposition scales and discussed with me privately (in Moscow - 1997) how to use these (real natural) semantic constructs to build the 117 dimensions of a semantic period table. In my humble opinion, knowledge science will be founded on this periodic table in the same way that chemistry was founded on the periodic table of atoms.

Thus a concept space need only to be defined as a set of locations (concepts) A = { a₁, a₂, a₃, . . . , a _n } where n is perhaps 1,500 (in most universes of discourses) and a set of "relationships" R { r₁, r₂, . . . , r_m } where m = 117!

The combinatorics are still large and create NP complete problems. However two things (somewhat related) come to our aid. First, the invention of Bjorn eliminates the NP complete problem in a specific fashion. Second a description enumeration of a hierarchical decomposition of rough (overlapping) categories of relationships and concepts produces a new formalism for navigation within the combinatorically reduced space of the virtual graph database. The voting procedure is a public domain algorithm that enables a simple routing and retrieval process:

I invented the voting procedure as a simplification of Russian quasi axiomatic theory:

Whereas the description of A can occur in any scatter gather processes. The description of R must be done with a non-probabilistic viewpoint regarding plausible reasoning (not just reliable reasoning). The method of cognitive graphics (Zenkin and Prueitt) can be used to vet visual acuity about meaning into a selection of some subset S of R so that the basic element of a concept graph in the form

< a, r, b >

where a, b are in A and r is in S, can replace the relational database concept of stored procedure.

This new class of virtual graph database stored procedures is the formal structure of the knowledge artifacts of the knowledge age (my opinion).