The material in this document is part of a series of provision patents filed by OntologyStream. The material is published for public and scholarly debate and to assist Ontologystream Inc in developing the prior art and will be referenced in the patent application to be filed sometime in 2004. Other material form the provision patent is at link.
From Paul Prueitt 12/19/2003 9:38 AM
The automated update of ORB
and retrieval index 
The name Knowledge Micro Pearls (KMP) is an interesting name.
We are all working on different things. Perhaps KMPs is a language term that KM ers will enjoy using as a means to talk about ORBs.
The automated update of ORB and retrieval index
But we have to find the January budget, somewhere. I think to myself, if somehow I could just get a budget for one more month that I could focus on the computer science with Nathan so that we could walk from the II core engine SDK completion into the system that integrates the SLIP , the ORBs and II into a single bundle.
The product could be shown and discussed at the NIST Friends of the Intelligence Community meeting in mid January and at my talk on the 28th at NASA Goddard.
We could establish the ORB retrieval and automated update as a visual interface into one or more “community bead games”.
The InOrb product will be able to
1) Automatically wake up each morning, or each five mins, and redevelop the ORBs and the Index for a community discussion that is either in a e-forum like
http://bcngroup.homelinux.org/forums/
or a html based hyper text document like:
http://www.bcngroup.org/python3/home.htm
2) The update of the index is easy because of the specified properties of our II core indexing engine. Once the II SDK is completed (January 2ed, 2004) this automation code will take three days of work to finish.
3) The redevelopment of the ORBs involves two parts
3.1) The inspection of the terminology usage. Have new words been introduced? If so what is the ORB neighborhood of the new word, ie what other significant words are co-occurring within a distance of one sentence or three sentences or a wider "scope" such as a paragraph?
3.1.1) The ORBs can be developed on the new materials and then new information added to the old ORB developed on previous material because the process is associative and communitive (means that an update can be done either way and produce the same result.)
3.2) The inspection of new co-occurrence involving old terminology. Again, an
< a, r, b >
element is developed exactly and only when the word a and the word b co-occur in the new forum material. So the relationship is structural. The relationship is then reified, to be given the proper semantics (via interpretation and cognitive acuity).
3.2.1) Our research group has a research resource that is donated from SchemaLogic Inc. This is a schema server that will allow a distributed community of practice to maintain a controlled vocabulary. It is a small matter really – but if not taken care of the knowledge management aspect of the over all system fails.
3.2.2) In our case this controlled vocabulary will be the set of significant words that are organized and shown as ORB neighborhoods.
3.2.3) It does not matter if there are more than one occurrence of an <a,r,b> element . The “category” of all occurrences is indicated by a single <a,r,b> element. There is a collapse of the subject occurnece into a single “subject indicator”. A neighborhood such as in Figure 1 is composed of and equivalent to a set of <a,r,b> categories.
Figure 1: The FFC (1997 – 2003) ORB neighborhood for “argues”
The structure in Figure 1 would be updated if the new material posted into the FCC public repository contained a co-occurrence of the word “agues” with any word (deemed “significant” by a community of editors, or librarians, communicating about the FCC public documents using the SchemaLogic schema (controlled vocabulary) distributed KM software.
Citizen centric groups and corporations will not
have to depend on the FCC, or other agencies, to supply subject matter
taxonomy or subject metadata. The
ORB development and indexed retrieval will be done from outside the agency by
for-profit, or not, vertical consulting groups. Each consulting group, or public interest group will have its own
set of significant words and Upper Fixed Taxonomy.
The Figure 1 graph is connected to the larger ORB graph via a link from “involvement” to some other significant words, a link from “commenter” to other significant words and a link from “against” to other significant words. The set of <a,r,b> elements associated with the graph neighborhood in Figure 1 is :
{ <center, r, against>, <center, r,
unpersuasive>, <center, r, notic>, <center, r, involvement>,
<center, r, diversified>, <center, r,demonstration>,
<center, r, conversely>, <center, r, commenter> }
The ORB has an In-memory “key-less” hash table (called an I-RIB) where the <a,r,b> elements are stored. This table allows for updates, deletes and addition and when changes are made the visualization automatically reflects this update.
Business Model
The targeted business model is in support of a many-to-many communication system that enables/instruments the social discussion of many (tens, hundreds, or millions) of individuals to share in an updated ontological representation of that social discussion.
This visual interface into the world’s social dialog may be more interesting than Internet porn, television, or on-line gambling; and may be coupled with on-line role playing games. The system can be developed easily over the next two months, see the InOrb investment plan.
The ORB (Ontology Referential Base) construction acts as a type of poll, in that individual contributions to the social discourse is not initially important, only the over all structure of themes. The automation of ORB development is supplemented by the hand of bead masters, following the book written by Herman Hesse, The Glass bead Game. Herman Hesse’s book is published in almost all human languages and is read by most of the world’s knowledge scientists (we estimate that about 600 knowledge scientists exists today).
The ORBs can be represented in a language neutral form. In fact the ORBs can be developed form astrophysics data sources, or any source of data from physical of linguistic measurement instruments. The principle has been demonstrated by knowledge scientist, Don E. Mitchell, in the domain of voice recognition technology for medical information acquisition.
When coupled with the Human Markup Language standard and the Topic Map standard, this ontological representation of structure can be applied to increasing the correctness of intelligence (business intelligence or national intelligence). So this vertical application area is one possible source of income for the research group that developed the ORB technology. Our efforts at getting the attention of In-Q-Tel and Army Intelligence (General Alexander), continue.
Business intelligence leads naturally into an analysis of hidden value chains and production/supply chains. The capability of the ORBs to representation structure in data sources is not limited to natural language, and thus will be applied to bio-informatics. This application to pharmaceutical research is one possible source of income for the research group that developed the ORB technology. The application to modeling medical and biological research is another possible source of income for the ORB patent owner.
The corporation dataRenewal Inc (Virginia 2003) was created to apply InOrb Technologies to a critical problem of social interest. The web site at dataRenewal is evolving to demonstrate the power of the InOrb technologies. We have already made the public document repositories of the FCC via subject matter taxonomy. As we have time, other public repositories are to be opened to citizen access.
The InOrb Technologies web site is at
An immediate possible source of income is from BCNGroup membership and gifts from wealthy individuals who see the social value of our work.
*************************************************
It takes a Community to Create a National Project
*************************************************
***********************************
Dr. Paul S. Prueitt
Research Professor
The George Washington
University
Founder and Director,
(1992) BCNGroup.org
Founder, (2000)
OntologyStream Inc
Knowledge Scientist
Cell: 703-981-2676
paul@ontologystream.com
***********************************