[138] home [140]

Friday, September 24, 2004

The BCNGroup Beadgames

Background discussions on a proposed

Anticipatory Technology Challenge Problem

Center of Excellence Proposal

Challenge Problem

National Project

White Paper on Incident Information Orb Architecture (IIOA) à

Orb Platform for Managing Incident Information

Abstract

The Orb construction provides a platform for ALL data mining and categorization of structure and function from the variable measurement of evidence produced by computers about events that have both a set of computer signatures and a real world circumstance.

This means that any data mining procedure currently done can be reproduced exactly using the simple Orb construction and convolution operations defined over sets of these constructions. The Orb equivalent will operate with

• one to three orders of magnitude greater speed,
• one to two orders of magnitude less memory requirements,
• has less entanglement with current commercial software systems and
• no compatibility issues.

Simple extraction programs can strip from any relational database or XML document information and encode all of this information into simple Orb standard constructions, having the form of an ordered triple < a, r, b>. The information is represent as sets. Convolution theory and very simple computer science acts on these sets. The convolution is performed on single pass over data in a set. The mathematical foundation to transformation of data is elementary to the calculus and to elements of real analysis, and has both an continuum and discrete version. In the discrete version, transformational grammars can be encoded as convolutions and thus structural manipulations of data can occur nearly instantaneously.

The Orb construction and transformational grammars also provide a complete and independent platform for the scientific, empirical plus theoretical, study of patterns of computer data signatures and functions that these patterns have and can play. Anticipation of intended function can be measured and alerts sent when circumstances appear to warrant this. Anticipatory science can be developed situational to an emerging threat.

The use of the Orb Technology is portable and can be used in autonomous mode. A new Object of Investigation can be defined by anyone, including non-computer scientists. A public methodology will be available to guide a new user through the entire process of discovery regarding patterns of signatures and structure to function observations.

Properties:

Real time acquisition and structuring of data is assumed to be the highest priority. This type of use is supported by the following properties.

• Orb technology bypasses intractable problems in Artificial Intelligence by bringing the informed Human in the Loop to bear on questions of judgment.
• Human-centric Information Production (HIP) is enabled with full local, unencumbered, control over software functionality.
• Absolute Global interoperability is based on non-proprietary set and graph theory.
• The Orb software is platform independent and can be implemented with a very small operating system.
• The Orb technology with a minimal Linux distribution can be fit into a hand held computer.
• Orb Notational System has been publicly available since 2003 and is being extended as part of a scholarly literature.
• Orb data encoding paradigm makes a correspondence between minimal computer encoding and fetch -execute cycles and the Orb Notational System.
• No proprietary anything, so the Orb technology can be understood without NDAs.
• High levels of security can be easily added based on simple encryption of network communications between Orb Platforms.

Present applications include the SLIP browsers and Orb Suite. Both are available in the public domain and as used in training.

Past experiences (to be detailed on request)

The Team : Dean Rich, Nathan Einwechter. Paul Prueitt, Peter Stephenson, others (to be identified)