Dark Triples

Jim Hendler has sent a message [1] to the sw coordingation group which states:

[[
 >
 >If the WebOnt language is to
 >   1/ use RDF syntax, and
 >   2/ be an extension of RDF, and
 >   3/ to be able to entail appropriate class memberships,
 >then the WebOnt language needs to have a facility for having RDF triples
 >that do not denote in the RDF model theory (a.k.a. dark triples) and URI
 >references that do not denote in the RDF model theory (e.g., URI references
 >that only exist in dark triples).


The WOWG has reached strong consensus on the above -- we have
determined that the 3 features above are mandated by our charter, and
that we cannot have all three unless RDF core does something with
respect to dark triples (as discussed above) We have several
volunteers from WebOnt WG who will be willing to engage with RDF Core
members in resolving this issue to the satisfaction of both groups
(this includes several people who are in both groups).
]]

I take that view that we are all in this together and if webont have 
reached strong consensus on this issue, this is something RDFCore should 
take seriously.  Consensus is often hard fought and is thus valuable.  I 
don't think that behaviour that might be construed as confrontational is 
helpful.

We asked Jeremy to post a motivating example of the problem and at least to 
the best of my understanding, we have had three different responses:

  Jeremy: students and employees
  Pat:    no example, denying students and employees
  Jos:    Peter F. Patel-Schneider's encoding of Russel's paradox in DAML+OIL.

I'm not sure how Pat's description of the problem relates to the PPS 
paradox - the paradox may be an example of Pat's general description.

I understand that the PPS paradox, as currently formulated, is no longer an 
issue as some of the constructs used in it have been withdrawn from owl.  I 
don't know whether that the solves the problem or whether it can be 
reconstituted in another form.

What is apparent, is there is no consensus on a description of the problem 
that dark triples will fix and how they will fix it.  So the first step is, 
I suggest, for the folks who have been involved in the discussion so far to 
agree at least on that.  I would expect such descriptions to include both 
an abstract, but accessible description of the problem and at least one 
concrete example.

Whilst we are aiming for last call is not the most welcome time to receive 
new issues.  However, I suggest we should work constructively with webont 
to assess this issue and decide how best to deal with it.  Questions of 
charter will be for the CG; it may be best to deal with it now, it may be 
best to handle it in a new spec synchronized with the webont process.  To 
reach an informed consensus on that question, we all need a clearer 
understanding of what the problem is, where we are in understanding the 
possible solutions and how those solutions relate to other issues we have.

Before we start forming our judgements, lets get the facts sorted out.

Brian

[1] http://lists.w3.org/Archives/Member/w3c-semweb-cg/2002Apr/0010.html

Received on Thursday, 18 April 2002 11:39:56 UTC