Monday, December 12, 2005
The issues related to having a common, widely
used, and available ontology
I hope that each person does not mind my many emails. I suspect that the volume of email will go down soon as there seems to be some resolutions taking place in regards to some future work. Please just ask if this it too much trouble.
I mentioned in a recent note the concept of a semantic cover as being a set of concepts that are enumerated, having a near axiomatic nature in the sense of (as a matter of human interpretation) being relatively complete for purposes of talking about a specific something of interest, and of being relatively independent so that a minimal number of concepts are used. This notion of semantic cover was one that developed in 2000, for me, as a method of Descriptive Enumeration as discussed on slides 6 and 7 at:
http://www.bcngroup.org/area3/pprueitt/private/KM_files/frame.htm
I do understand that simple enumeration methods are commonplace.
I suggest to each of you that some closure might be made on the issue being discussed in the ONTAG Working Group.
a sample of the ONTAG discussion from my point of view is at
http://www.ontologystream.com/beads/nationalDebate/188.htm through
http://www.ontologystream.com/beads/nationalDebate/280.htm
The purpose of this ONTAG group is to develop some common understanding of how to make common, widely used and available ontology (of some type). However, the point of this email and the request for position papers is (we conjecture) that this ONTAG group is absolutely dominated by a specific viewpoint - a viewpoint which is steeped in a paradigm which is not recognized as being in fact narrow and largely on the way to being discounted (as AI now largely has been - at least). So even the suggestion that there is something that needed to be de-constructed about the notions of logic, inference and induction is considered to be a suggestion that should be banned.
The Mitre report:
http://www.mitre.org/work/tech_papers/tech_papers_04/04_0603/04_1175.pdf
"Towards the Use of an Upper Ontology for US Government and US Military Domains, an evaluation" came out in 2004, and so is dated. However, the ONTAG Working Group seems to have not progressed beyond the report (which as it was presented to the group in email seemed to me to "just be coming out now").
I note that the central part of the Mitre report talked about types of ontology in such a way as to make me feel deeply marginalized - simply because the discussion defined away any possibility of talking about MY type of ontology, or Tom Adi's ontology or even Richard Ballard's ontology. The report almost leaves no opening for an examination of the issue of interpretation and degeneracy (as discussed by Edelman).
So back to this notion of descriptive enumeration and the related notion of semantic cover.
I conjecture, that if each of you were to provide a focused well written and objective viewpoint about SUMO (Suggested Upper Merged Ontology) that the collection of these viewpoints would cover mostly an alternative paradigm to that which is implicit in the Mitre report. John Sowa is able to see both viewpoints and perhaps might honor us with some objective analysis about my requests.
Is it possible to make an actually breakthrough in the knowledge sciences, given the existing funding patterns, John?
I recognize that there is a great deal of work that would have to be done for Paul Werbos, for example, to become familiar with the concepts, organizational methodology and implicit assumptions related to the SUMO. So he will have to view the whole notion of W3C type ontologies and perhaps compare them with OASOS viewpoints or viewpoints that come out of an experience with neural biology. I am hoping that my conversation with him, and with each of you individually over the years will define an on line volume and then a actual face to face conference (perhaps here in Taos New Mexico).
If this request is honored, we might be able to produce a collaborative document over a period of a couple of weeks or less... that brings attention to the fact that the Mitre paper about concept representation mediation of human interaction during critical response activities is NOT fully represented by the Mitre paper.
At least this is what I hope each of you might agree on, to some extent. That the Mitre report does not go far enough to represent alternative viewpoints about how concept representations (or even semiotic systems - as defined by Ricardo Gudwin for example) might be used in mission critical military or social crisis response.
My hope is that an on line volume might be produced - with specific papers written by ten to fifteen individuals. One last thing.
You all know me in some way. I have struggled so hard on this issue as to have literally destroyed my private life ... and am now trying to make a new life based on lessons learned. I truly believe that the nation and our people do not have the tools needed to address crisis management.
The project that I carefully set out on now is designed to present professional attire while advancing an deeply informed awareness of the issues related to the Semantic Web.
Paul Prueitt
Taos Institute