Monday, November 08, 2004
Center of Excellence Proposal
à
White Paper on Incident Information Orb
Architecture (IIOA) à
Adi Structural Ontology Part I
à
Cubicon language descriptive
à
On methods for the development of ontology
On the Ontologies needed for container monitoring
On the nature of the AI polemic
Mills Davis
Top Quadrant Inc
Mills,
Thank your for the excellent phone conversation this morning.
This note is communicated to those who have received glass bead game communications in the past, and who I feel should continue to understand the background processes related to the planning for the National Project to Establish the Knowledge Sciences.
Please share this with Irene and Ralph and anyone who you feel should be aware of the developments that are indicated in this communication.
I include Dr David Day at Mitre because of the BCNGroup proposal on anticipatory technology and ARDA's non-interest in our approach.
http://www.ontologystream.com/beads/nationalDebate/challengeProblem.htm
http://www.ontologystream.com/beads/nationalDebate/146.htm
I have not as yet received any information about the proposals that have been accepted, nor about the comments of the review panel.
I have included Dick and Rick because they will feel a need to address what is not said.
Rick has a seat at the table of individuals considering the large scale and deep implementation of ontology in the military and at the State Department. He may have something to say about the general problem that the BCNGroup founding committee has referred to as the AI polemic (see references below).
Specifically Dick Ballard's Knowledge Foundations software system called Mark 3 could easily be the system adopted by DHS for all shipping container monitoring.
But, in my opinion the Mark 3 is not completed. Dick agrees that he is perhaps 6 months and 1 M away from completion of his ontology builder software interface and the encoding technology underlying the Mark 3. If the Mark 3 were completed, and was not burdened by proprietary hiding of how it works, then the Mark 3 encoding of knowledge into n-aries which reference the precise details about the following would be (very close to) optimal.
Ontologies are needed for the following:
1) physical resources related to and used by processes that monitor, or
that will monitor, shipping containers.
2) physical resources related to and used by processes that make
responses to crisis involving the material contents in, or likely to be in,
shipping containers.
3) the real time human information traffic arising from normal
operation of monitoring and response mechanisms
4) the real time human information traffic arising from operation of
monitoring and response mechanisms during critical incident response.
5) repository of normal shipping behaviors
6) repository of monitoring and response behaviors
On #3 and #4 Peter Stephenson (Eastern Michigan Univ) and I have a White Paper:
I propose to advance the resumes of two individuals who have RDF and OWL or Protégée experience and experience with the knowledge management / web services standards as understood by Irene and Ralph.
The strategy is that the government client, would be lead to the more advanced Top Quadrant Inc type notions of how ontology is created and used. This would occur not from a direct proposal for architecture, which I would be able to do, but through tricking the Prime into using Newcomb and Biezunski's complete notion of a topic map (in the later Wittgenstein sense). We would, in a way that is not expected by the client or the prime, deliver what the BCNGroup community has talked about as being a "Knowledge Operating System" with usable interfaces between a descriptive ontology, having minimal logical mechanisms, and human cognitive capabilities.
I specifically reference Leo Obrst's (Mitre) Chapter (Chp 7) in Jack Parks (Ed) "XML Topic Maps" as being a polemic. By this, I mean a definition of something, ontology in this case, as being precisely as defined by the polemic and not in any way allowing an alternative view to be legitimate.
I have discussed this notion of polemic in a recently revised preface to my on line book
The 2004 Presidential election brought home to me the notion of a polemic. Senator Kerry was "created" by the Bush campaign to be someone who could be defeated based on misperceptions of who he is. Perhaps this can be justified based on the need to have this unambiguous leadership. But the fact that a false understanding was created is a matter of observational fact. One might say that the justification for the invasion of Iraq was a similar polemic.
One should expect additional uses of polemics represented as truth that cannot be doubted. And perhaps this is acceptable given the actual threat our civilization faces in regards to the use of WMDs by terrorist, and delivered via the shipping system. I did vote for Senator Kerry, as many Americans did, but I feel safer with President Bush being in office for 4 more years.
The issue concerning ontology development and deployment however is about polemics that are created for the purpose of controlling the social discourse.
My daughter Jenni has been writing about the polemics developed in the West about the nature of Islam, and Tom Adi and I have been talking about the use of Orb and Readware real time representation of concepts in text as a means to develop many cultural viewpoints as part of a BCNGroup Glass Bead Game.
In any case, the polemic that is rooted in the traditions of AI, and is now re-expressing as Leo's polemic description of Semantic Web, is a critical issue. This critical issue is related to our community’s interest in showing that Semantic Web advocates needs to understand more than they have.
All members of the Semantic Web community need, in our opinion, to more deeply understand the interface between humans in everyday life and the encoding of information in the form of n-aries.
N-ary encoding is simple. The power comes form a question in a question and answer interface to the measured structure as descriptively encoded into a set of n-aries:
{ < r, a(1), a(2) , . . . , a(n(q)) > }
where r is a relational variable (if fixed to be "co-occurrence" or "is resource of" then we have SLIP Orbs or RDF) and a(i) is an enumeration of topics.
n(q) depends on the n-ary, but is most often in real life less then 100.
I will not go further here because there are too many things to say about frameworks.
The development of ontologies of this nature can be done in RDF for each of the 6 types of ontologies needed (as listed above). However, the interface to humans in real time has to be agile and not controlling of the humans who will use these ontologies in times of crisis.
A nuclear weapon has just exploded with no warning in a city in the Mid West and all shipping has to come to a stop so that the possible second and third explosions can be mediated.
The bottom line is that neither the Prime contractor nor the client understands anything about the history that I am referring to in my presentation of the foundations of a future science of knowledge systems. They do not know very much about the alternative history that Leo Obrst has created in his chapter, and in his many presentations.
What the client does know is that there is a long history of failures of the systems that Leo talks about with deserved pride. These are hard problems, particular when approached as if the computer can be brought to “understand” and make inferences.
Jim Hendler would say that lessons have been learned and that the Semantic Web community (as he and Leo understand this community's work) is ready to try again.
I agree that Semantic Web technologies will be funded again.
However, the flaw that is persistent in Hendler and Obrst's presentations; has not been understood nor accommodated intellectually by them. They simply will not look at the neuroscience of the systems theory that I have placed in front of them.
To quote Obrst:
"The primary advances in knowledge representation over the last 15
years can be attributed to the increasing role of formal logic in
representation and reasoning formalisms and to the development of theories
based on logical and mathematical principles."
I must apologize for the state of the chapters of the on-line book, but I address the problem of the limitations on formalism as a tool in Chapter two.
The arguments I bring forward here are simply not arguments that Leo or Jim will address, because these arguments pulls the rug out from under the notion that mathematics is the path to forming proper ontology. Both have told me that they are unwilling to talk about this argument.