Friday, September 24, 2004
Background discussions on a proposed
Anticipatory Technology Challenge Problem
< This is a note to several of the computer science department
faculty at George Washington University. It is posted here so that the communications can be referenced. The discussion is part of a number of separate
discussion we are having with faculty and administrators in five universities, concerning planning for the
National Project. >
Simon
Berkovich,
Dear old friend,
We have known each other for 12 years now, since my
conferences at Georgetown in 1992.
I seem to be in a procedural loop with the Chair and OSP
at GWU. As I have been in this place
with the department before, I will make a different response than in the past.
The NSA and other agencies of the IC are interested in
technical solutions to the problem of knowledge generation, knowledge vetting
and the measurement of informational structure in the Internet. Many people feel that my group has a new set
of innovations that offer new capabilities.
But we need help, both financially and institutionally.
What we find is that everyone is busy with something
else. This is understandable, and is
consistent with the fact that there are sets of problems, in intelligence
production, that are currently intractable.
So the proposed Manhattan-type Program to create a new
type of Information Science has a clear context. Some thing unexpected needs to occur. Anticipatory Technology may be the answer.
In the attached Center of Excellence draft proposal (.::.),
I clearly communicate a Challenge Problem.
I am proposing this Challenge Problem to Dr Day at Mitre (acting for
ARDA) as interdisciplinary, and yet firmly grounded in computer science.
I also clearly reference a body of scholarship in regards
to why some of us believe that computer science has "left the
barn".
I clearly state the proposition that our society is not
prepared to understand the issues related to memetic mapping ongoing by the
intelligence community and by corporations involved in targeted advertising and
hidden methods designed to invade personal privacy.
I state clearly the proposition that the mathematics and
computer science undergraduate curricula are missing essential components,
required IF our citizens are to have the preparation required in an age where a
War on Terrorism might last decades!
I propose a National Project to create a K-12 curriculum
in the knowledge sciences. What could
be clearer?
The date in which GWU's OSP would need to consider my
existing and completed proposal, with specifics and a budget, is in
November. The Oct 13th date is to
establish a challenge problem related to anticipatory human computer
interfaces.
The "opportunity" for the department is to open
a discussion between a group of technologists and scholars, listed in the
proposal, the GWU CS faculty and students.
The Center of Excellence would create a distributed open collaborative
Glass Bead Game:
http://www.bcngroup.org/area2/KSF/nationalProject.htm
What could be more relevant?
In response to a published RFP, I am proposing as one of
the 2005 ARDA challenge problems, a substantial "interdisciplinary"
investigation as to what "anticipation" is.
Other disciplines include:
Human Mark-up Language standard,
Topic Maps standard,
abstract intelligent agents,
data structures,
pure mathematics,
mathematical biology,
topological logics,
ecological psychology,
evolutionary psychology,
general systems theory,
complexity theory,
graphical representation of human concepts,
natural language parsing,
social network theory,
knowledge and innovation,
political science,
diplomacy,
biologically feasible mathematical models of human brain function,
social network measurement and modeling, and
models of deep structure in natural language expressions.
see: http://www.ontologystream.com/beads/nationalDebate/115.htm
The computer science issues are clearly stated in my 10
page challenge problem proposal, attached:
1) in what way could data mining produce computational resources that
serve as a cognitive primer via visualization.
2) can cognitive graphs, using John Sowa's term, be used as a
"computational element" in the production of cognitive priming
leading to anticipatory responses
3) is the formalism based on n-aries, to use Richard Ballard's term, a
sufficient and proper formalism, both for encoding structure of co-occurrence
of n-grams, terms, and/or phrases; or for encoding knowledge representations
directly (as advocated by Dr Ballard).
4) can algorithms generally classified as AI algorithms, such as latent
semantic indexing and associative neural networks, be used in a weak form as a
means to evoke in humans an anticipation of "what is next".
5) is there a VLSI encoding of my Orb construction
that allows a very fast and scalable measuring of invariance in data flows?
My ten-page write up certainly suggests some
possibilities. (.::.)
One might think that these possibilities could be
presented to the faculty and students in such a fashion as to evaluate the
merits and to consider contributions that could come from the faculty and
students.
What I propose is to make an announcement to the
department that the problem of "anticipatory computer technology" was
being considered. I am inclined to not
make GWU the prime on any contract to ARDA, but to specify a subcontract
POSSIBILITY to support
1) the collaboration with one of the Faculty on the
data structure theory aspects of the Orb encoding.
2) the support of one graduate student.
There
is simply no requirement that OSP sign off on anything unless my corporation
becomes Prime to this $900,000 contract.
I will not use GWU name, and will simply state that x amount of money
will to provided to support y and z subcontracts to a willing university. So no one need to bother unless the funding
is actually in my control.
My sight is on the Center of Excellence proposal, and
though there are current discussions with Senate and House staff, this proposal
will not seek funding until 2006.
I expect to visit with members of the Administration at
GWU and at Georgetown, to determine if an interdisciplinary group of faculty
will engage in the issues related to the Proposed Center of Excellence and the
proposed National Project.
The department's support has been deeply appreciated. Please consider establishing a time for a
seminar so that I might make the faculty and students aware of the nature of
the Orb (Ontology referential Base):
Paul Prueitt
Research Professor
The George Washington University