(soon)
Communicated by Paul Prueitt: 12/23/2003 9:07 AM
The ability to find the right next element in a curriculum is essential to optimal educational experience. This optimality now only occurs in rare situations.
The selection of the next curricular element to use in distance learning, or face-to-face learning, is similar to the ability to find the second word in a subject matter retrieval using the standard key word search.
What the new InOrb technology offers is a type of personal, or community, knowledge management system based on facilitating the personal knowledge of how natural language is used (in human communication systems).
I have addressed this in the new paper at:
http://www.bcngroup.org/python3/thirtyseven.htm
In this paper we develop an architecture that used an inverted indexing engine, InOrb ORB and SLIP technologies go together to provide a completely new capability for retrieval.
As you read in
http://www.bcngroup.org/python3/thirtyseven.htm#_Basic_notation_and
the complex transmission of "Ontology Referential Bases" (ORBs) is formally set up to allow a type of pattern completion using the voting procedure, invented in 1995. The result is no longer a search process. It is both a retrieval and inference combined if the inference information is present.
Because the ORBs are visually represented as very simple graph structure, human cognitive acuity can provide the inference, or make correction to inference developed by the ORB space. These are easy technologies to use and they will be free to use during the nine months covered by the dataRenewal LLP investment plan. (ask for copy )
This investment plan is but one of several that we offer as a means to combine the work that has been done with a small amount of capital needed to provide the new technology to the public.
The reason why this ORB retrieval/inference works is due to the nature of physical and energy distribution in a wavelength spectrum. How biological systems work, including how human cognition works is dependent on underlying physics.
The abstraction that is used to create computer programs and structural encoding of the linguistic variation happens to match this natural physical phenomenon. The “knowledge science” is correct, and therefore the knowledge technology is very powerful, simple, and quite different from the Information Technology that our society is now spends about 2 trillion dollars per year on.
We, a group of primary inventors and scientists, can create and make available an infrastructure of educational processes and an economic system based on internal accounting software.
If the new software is embedded in the Knowledge Sharing Foundation concept, because of a scale of use, the capability to retrieve knowledge evoking information becomes essentially un-encumbered by a use charge - like telephones, we just use them.
Other tools will be added to provide just in time information having very high quality fidelity (truthfulness). In each case, however, the scale of use can provide to the owners of the underlying technologies considerable wealth.
Society can go on to do other things, like bio-defense systems and the reduction of poverty.
Python3/Linux Knowledge Technology Toolkit for Kids (K-12) CD
The new ORB retrieval is ultimately very simple to do and to explain. New types of jobs will be created, and an economic boom will follow as the Knowledge Technologies Sector Ignites. The benefits are to society, to the inventors and owners, and to the educational process that is needed if the knowledge sciences are to become part of our school curriculum.
ORB retrieval/inference is retrieval, and pattern completion, without search and follows the same architecture that some of the artificial neural network architectures follow (for example those based on Grossberg's work) where functional architectures and metabolic processes are those actually observed in the brain.
These systems, both the mathematical ones and the natural ones, provide selective attention and what some call a “negative search”. In a negative search, the retrieval occurs directly without search but with an impedance metric. So the direct consequence, the thing perceived in the vision system or the documents retrieved in the document management system, can be immediately validated by human awareness.
If there is a mismatch the mismatched "signal" has a depleted ability to play in the next direct retrieval and so the next retrieval cycle produces something different that then previous cycle. This cyclic dependency on a reality check follows our understanding of work in the cognitive neurosciences. The direct retrieval occurs because the retrieval structure already is in the right place in a spectrum. The retrieval structure and the elements to be retrieved are structurally the same.
It is possible that language is used in an unusual way to disguise the meaning or to rely on tacit knowledge possessed by the listener. But there is still structure in the co-occurrence patterns.
All text understanding systems have to rely on this structure on co-occurrence of words. See our work on Latent Semantic Indexing and the ontology lens.
This artificial neural network architecture follows work that Dan Levine and I started in the late 1980s.
http://www.bcngroup.org/Macrocognition/two.htm
Having the Ontology Referential Bases with local neighborhoods oriented around the elements of a community-agreed-on fixed upper ontology
http://www.ontologystream.com/beads/enumeration/taxonomyNote1.htm#_Two_level_Fixed
allows the Instant Index engine, which is quite different internally from Google, to immediately go to that exact part of a very small bit map that is the subset of a standard inverted index. An immediate and direct retrieval returns the document sections that have exactly and only the co-occurrence of words specified from the ORB neighborhood (seen as the blue simple graphs).
http://www.datarenewal.com/services.html
From the above URL, look at one of the advanced search (fables or FCC). The retrieval is immediate, complete and has 100% precision. The precision recall is perfect because what it is retrieving is all and exactly those documents with the co-occurrence of words that are specified. The semantic fidelity to what is needed, or what is sought is high, because of the way natural language is used within communities. The exact knowledge of what words are co-occurring is provided by the InORB technology in the Subject Matter Indicator neighborhoods.
The semantics is not addressed except as imposed by the community (in declaring controlled vocabularies), or in the interpretation that happens so naturally and with such little effort by the human.