Wednesday, May 11, 2005
concept of standardizing unique definitions for words
Eric Miller
W3C
Eric,
I just noticed your name on
http://labs.intellidimension.com/uniprot/default.rsp (and your email address) I have wanted to continue our discussion.
The BCNGroup is working with Geoff Chappell and will have Intellidimension’s RDF repository and inference engine under the Global Information Framework..
I deeply appreciated the private discussion we had, at SWANS, on the importance of human centric ness to the semantic web.
I am wondering, Eric, if you have thought about the concept of standardizing unique definitions to words as being “equal to” a specific meaning. I am sure you have, but I wonder about how you come to reconcile the terminological drift … one sees in natural language use.
I know that this is one or two or three core principles that W3C has recommended standardization around. There is good reason for the recommendation that standard semantics be described (within namespaces). There is a strong relationship to the use of the URI/URL to place the “correct” semantics for a word at a specific place (within a namespace).
I am sure that there have been thousands of conversations about this. But I am not convinced that a balanced viewpoint has been available to senior W3C officials.
The Roadmap takes a different position, while acknowledging the importance of some key standard concepts… I think of the integers as being an example. This kind of upper level abstract ontology does do not change and has no contextualization that has to be accounted for.
See: http://www.ontologystream.com/gFT/5-modularOntology.ppt
Slide 2 and 3
I feel I am missing something here about what the W3C sees as the near future.
Paul Werbos and I recent had a conversation about the deeper issue of [64], [65] fundamental determinacy in conjecture by Wolfram on emergence….
With my pointing out that G Edelman’s concept of response degeneracy is an attempt to bring objective science to the kind of thing that goes on as emergence occurs. I would say that Wolfram is first school and Edelman is second school.
We would love to see how this kind of discussion might influence the work at the W3C.
Paul Werbos is one of the thought leaders in the neural network community and has seen this issue from his perspective as a leader.