Comment on DLP tractable fragment text (http://www.w3.org/Submission/owl11-tractable/#4)

Pardon me if this comment has been covered in previous dialog (please
point me to it if this is the case), but I have a comment about the
DLP tractable fragment in the OWL 1.1 tractable fragments document.
It states that the an OWL 1.1 reasoner would draw a different
conclusion than a LP reasoner would from a "DLP" ontology.  The DLP
definition it uses makes reference to literature that describes a
weakened form of Horn clauses that excludes the features it presents
as a cause for the discrepancy (negation as failure - so called
"closed world assumptions").

However, in the paper it makes reference to, the corresponding DL
(which is the domain of the translation) could not include assertions
that would cause LP reasoners to reach different conclusions to a
corresponding OWL 1.1 reasoner., since the translation is "sound and
complete" according to the cited text.

It seems some additional clarification is needed in the DLP tractable
fragment, around this subtle, but important point.

Received on Monday, 17 September 2007 03:20:17 UTC