Some summary comments on

Justification for a October Workshop (Computing & Cognition)

to

Dr. David Alberts, Director of Research CCRP

Office of the Sectary of Defense

( Index )

Dr. Paul S. Prueitt

Founder (1992) BCNGroup.org

Founder (2000) Ontology Stream Inc

Research Professor, George Washington University

Anthony Cordesman, Center of Strategic & International Studies is pointing out, to the House Intelligence Committee, that the intelligence technology and methods that we have are not consistent with the uncertainly and deception that characterizes the threats. For example, he says that in the state of the art today, there is no way that one can develop high fidelity machine-ontology about weapons of mass destruction.

David Alberts, Office of Secretary of Defense, is pointing out that a disruptive transformation of knowledge sharing capability is necessary. I am pointing out to Dr. Alberts that such transformation of capability requires a disruptive transformation of procurement processes.

This is the context of Ontology Stream’s current work. We are concerned with how does one develop the cognitive - social – computer science so that a new kind of capability is made available?

It is clear to me that IF a new capability is developed that we would have two new clients, OSD and DHS. But more importantly, we will have a capability that would allow third parties to develop a new market having an almost unlimited economic impact. The problem is that the procurement and investment processes have gatekeepers whose interests lie elsewhere.

The First Knowledge Sharing Core could produce a capability that quickly produces machine ontology that is immediately reifiable by collaborative process involving humans at several levels of constructions. This is NOT an Information Technology (IT), as defined by the current market, but is a social/cognitive science grounded Communication Technology (CT), relying directly on humans in real time.

Thus the new capability is not one that is derived from the false assumptions of the academic school of artificial intelligence, to include the swarm and intelligent agent theory. This new capability depends on the thoughtful use of linguistics and first order predicate logics, as well as a manipulation in real time of the underlying elements of this logic, and linguistic rules that can be flexibly managed by normal people.

One key to understanding how the new capability is to be stood up is to review the architecture and software features of the SchemaLogic SchemaServer. The key to this system depends on a localization of terminology used by more than more community into “containers” where reconciliation processes evoke from human stakeholders rules over how terminology, generally in conflict, is managed when the community members are in dynamic, mission critical, discourse.

SchemaLogic Inc made several attempts to interest the federal government in the SchemaServer taxonomy management system, but found the entry to this market absolutely controlled by large incumbent vendors and consultant corporations.

Similar keys to the First Knowledge Sharing Core are found with the programming language for Natural Language Processing (called NLP++) developed by a small company (of two individuals) called Text Analysis International Corporation. TAI founder Amnon Myers has focused for several decades on the problem related to programming rules that allow the analysis of linguistic structure to be agile and “evolve” into a domain specific context, rapidly. Ontology Stream Inc offered this work, in 2002, to primary consulting and vendor corporations whose rich funding sources have been used to dig holes in the Earth and build expensive building, only to realize that the rent for the new buildings was too high to occupy. Moving the location was then the source of additional funding.

Recommind Inc has a new mathematical technique based on the best work done during past decades in stochastic and probabilistic work related to the examination of case grammar, and generalized case grammar. In a case history that examines Recommind’s initial attempts to market to the intelligence community, one finds a story very similar to that told by SchemaLogic’s Co-Founder, Breanna Anderson.

CoreSystem’s founder Sandy Klausner has developed over a period of twenty years a methodology and capability for designing computer processes using an iconic language that is superior to the NextStep Application Builder (first seen in 1988), and to UML type design methodology (available from Rational Rose). However, this methodology and capability is so disruptive in principle to .NET and to Java, that we were instructed, by consultants, NOT to expose anyone at DARPA to the excellent Macromedia presentations that objectively demonstrate the potential that would come from a complete re-factoring of all intelligence Information Technology knowledge sharing systems using this iconic language.

History is littered with mature text understanding capabilities that have been destroyed by one of two processes, often coupled together in a “unconscious” memetic collusion. The first process, in terms of influence, is the government’s long-term commitment and persistent funding of incorrect paradigms produced by the strong AI school. This funding has established an authority over both the government procurement of research and deployable Information Technology systems. The second process is that governed by a Venture Capital community whose personal wealth increases regardless of whether the technology start-ups fail or succeed in the commercial marketplace.

The procurement process is filled with illusions and deceit. It appears as if the best technical work, using the best innovations, and honoring the patents developed by individuals (such as Bjorn Gruenwald – Founder of Primentia Inc) is being sought, identified and procured. This is far from a reality shaped by greed and power.

A few, out of many, innovations are patented and made ready for market and are part of a procurement by the federal government. We can name Entrieva Inc and Applied Technical Services. But in each case, the underlying technology is revealed in a less than optimal fashion, when judged by objective science, and the funding made available to the primary innovators is often less that 5% of the total spent by the federal government on (directly) related evaluation and consulting services. These private interests have an interest in, and regularly receive, large personal bonuses.

An investigation reveals that these consultants do not have the technical expertise to make objective evaluations, and would conduct evaluations based only on locally defined business outcomes. Moreover, as a community these consultants do not represent a diversity of opinion that can be found within the community of scholars that we propose bringing into a series of workshops about these issues.

http://www.bcngroup.org/area2/KSF/KSFconference.htm