RE : epistemological level?

Hi Howard,

Sorry, I shouldn't have used a word such as "epistemological" here. I guess I had in mind something much more oriented towards Knowledge Representation(as a subfield of AI)-oriented [1] (it's not the original Brachman paper, but it sums it up quite well I think) than what you are thinking of.

Note that SKOS is Semantic Web-biased, which means in particular that we are quite reluctant to get involved with representaton mechanisms that require more than triples (and "Jack pushed Jill" will be enough for a number of applications, I believe).
Maybe that's too simple for your own needs. In that case please do not hesitate to read SKOS as "Simple Knowledge" Organization System!

Antoine

[1] http://www.loa-cnr.it/Papers/OntLev.pdf


-------- Message d'origine--------
De: public-esw-thes-request@w3.org de la part de Howard Burrows
Date: mar. 04/11/2008 14:28
À: public-esw-thes@w3.org
Objet : epistemological level?
 
Hi,

 

My question about SCOS (Simple Concept Organization Scheme) came from this
view: a system that purports to organize "knowledge" should help organize
answers.  It should help us choose between the applicability of expressed
statements.  

 

Antoine raises the issue with his use of the term "epistemological level".
Epistemological levels are not in the ontological narrower/broader,
class/subclass arena.  Epistemological levels occur in a ranked list of
types of justifications and warrants.

 

As a simplest example, consider the question:  "Did Jack push Jill, or did
Jill push Jack?"  

 

We can express the two answering statements with triples: "Jack pushed Jill"
and "Jill pushed Jack".  But it seems like we require something else
entirely to provide the supporting research that determines if either one is
true. 

 

Both a SCOS and a SKOS will need to handle the psychological, cultural, and
philosophical nuance noted in the earlier discussion.  However, if our
current scheme only helps us find ideas, it seems more like a "concept"
scheme and SCOS.  I don't think we get to call it a "knowledge" scheme until
we capture statements together with the structures that relate them to their
appropriate use.

 


Howard Burrows

Supporting Research

Durham, NH, USA