The Knowledge Sharing Foundation
The knowledge sharing foundation concept was first
developed (2003) as a suggestion supporting the US intelligence agency needs to
develop information about event structure.
Previous to this, a small group of scientists have talked about the need
for a curriculum, K 12 and college, to support an advancement of cultural understanding
of the complexity of natural science.
By natural sciences, we mean social and cognitive science in the context
of human communication.
The suggestion to support new intelligence
technology deployments is predicated on the intelligence communitys
responsible use and on the co-development of an open public understanding of
the technologies employed.
Ontologystream Inc has developed an (fairly
complete) understanding of the types of text understanding technologies
available within the intelligence community.
Questions
and Answers
September
16, 2003
Hyperlinks
(click below)
Q-1: What is needed to support
awareness of events in real time?
Q-2: What is needed to support
community use of analytic tools?
Q-3: What are the benefits to
Industry
Q-4: What are the Foundation Elements
Q-6: Why are educational processes
important?
Q-7: How does the software
compensation model work?
Q-8: How are test sets made available
to the competitive communities?
Q-1: What is needed to support Intelligence awareness of events in real
time?
A-1: The
structure of data produced from measurement.
Real
time web harvest of natural language discourse (Memetic measurement)
Global
measurement of reports from medical professionals (Genetic measurement)
Measurement
of social network relationships and dynamic boundaries of social systems
Cyber
intrusion instrumentation and analysis (Cyber measurement)
Measurement
of all NASA Earth Observation Data (see proposal to NASA)
A-2: New computer
science based on stratified theory
With
an aggregation of invariance in the structure of data.
With
a separation of statistical and categorical artifacts in correspondence with
human memory processes and human anticipatory responses
With
the production of just in time machine-ontology formation as cognitive
enhancement
With
the development of event templates indicating meaningful constructs
A-3: Educational processes that
allow users of an intelligence system to work within the limitations of machine
and artificial intelligence.
University course credit
Professional Accreditation
Q-2: What is needed to support Intelligence community use of analytic
tools?
A-1: A distributed
collaborative framework
Tools
expressed as un-encumbered capabilities
University
certified educational support on all tools
A-2: Tool
stability and tool interoperability
Separation
of all module services from vendor control, with the appropriate payment for
actual use of intellectual property
Open-results
competitive testing of all modules, with modules expressed as open source
software so that code can be seen and understood
Proposed
use of CoreTalk Macromedia presentation [Mac] [Windows] and the Hilbert Engine
A-3: Community
based compensation infrastructure
Commercial
rights are protected with copyright and patents
Use-compensation
based on software self-accounting to honor copyright and patents
Micro-transaction
accounting and payment for services embedded in each software component.
Q-3: What are the benefits to
Industry
A-1:
Establish coherence within the market
space
A-2: Advance
the state of the art for information generation systems and open new markets
A-3:
Establish a new basis for innovation
A-4:
Intellectual Property mapping and patent
evaluation will result in a reduction of uncertainty over ownership
Q-4: What are the Foundation Elements
A-1: Processes
Text
Transformed into Structured Data
Unsupervised
Pattern Mining
Supervised
Categorization
Situational
Logic Development
Logical
Inference (ΰ
see current discussion about
induction and deduction)
Procedure
Learning
Event
Detection from Data Invariance
Knowledge
Flow Mapping
Social
network and linguistic variation analysis
A-2: Subsystems
Single-algorithm
Analytic Servers
Multiple
User Domain
Ontology based Inference Engine
A-3: The Human
Element
Knowledge
Encoding and Propagation
Information
Visualization
Cognitive
Priming
Multi-modal
interaction
A-4:
Single-Algorithm Analytic Servers
Latent
Semantic Technology
Self-Organizing
Maps
Concept-Based
Document Indexing
Context-Free
Grammar Parsing
Clustering
Supervised
Text Classification
Evolutionary
Optimization
Associative
memory and top down expectation using neural networks
Social
network theory and analysis
Q-5: Examples of Innovation
A-1: Categorical
Abstraction
Invariance
in the data is used to construct situational logic
Continuum
mathematics methods are used to derive an "implicit ontology" from a
body of documents or other data sources
An
"explicit ontology" is provided by human beings, e.g. in the form of
categorized sentences, and then refined using iteration
Human
feedback and inference rules are used to further refine & process the
derived classifications
A-2: Event
Chemistry
A
technique for searching datasets for signs of real world events
Takes
abstract atoms of invariance observed in data, and forms interesting
combinations of them
Requires
a Human-in-the-loop cognitive acuity to provide interpretation of meaning
Works
naturally with the output of semi-supervised text classification, clustering
and categorization methodology
Fits
naturally with "chemical compound" metaphor, where a period table of
atomic elements are discovered and used in event detection
A-3: Referential Bases
Post relational
database technology, using new types of algorithms
(type:value) pair data
constructions encode localization of information without schema
(type:value) pair data
construction organizational processes has well delineated correspondence to
human memory and anticipation
Referential bases
support stratified processing so the ontology constructions can be formative
and situational
Q-6: Why are educational processes important?
A-1: The
systemic development of educational processes involves
the development of
consensus on what are the separated techniques in computational intelligence
The mapping of
scholarly literature helps in comprehensive mapping of patent disclosure and
copyright
A-2: As this consensus develops,
the description
of general systems theory, cognitive and social science is made available
within the academic community
a "liberal
arts" education in the knowledge sciences is made available to
intelligence analysts
Q-7: How does the software compensation model work?
A-1: Analytic
features are to be replicated from existing software and implemented as separated
components.
A
mapping of all software based innovation in the area of computational
intelligence is developed based on latent semantic technology indexing of
patents and copyrights
In
cases where the core technology has legitimate ownership, then licenses are
arranged
In
cases where the core technology is developed by the government then the core
engines are made public domain
Each
core technology component is rendered in binary with an internal accounting
module that reports usage as part of a knowledge flow mapping and use
compensation (when appropriate)
Q-8: How are test sets made available to the competitive communities?
A-1:
The system of core objects is open to
innovation.
Negotiations
to acquire a new innovation occur through Intellectual Property mapping
processes and comprehensive testing of object inherit capability
Innovations
targeted for acquisitions are studied in highly structured usability testing
that includes deep education in the innovations' inherit capabilities.
These
acquisition studies are conducted in the public view and are not governed by
commercial processes.
Knowledge
Sharing and SenseMaking diagram
www.ontologyStream.com
Contact:
Dr. Paul S. Prueitt
Research Professor
The George Washington
University
Founder: Ontology Stream Inc
Director: BCNGroup.org