Re: ISSUE-3 and RDF simple entailment

The aspect of Jeremy's message that troubles me (but was probably  
clear to many in the WG) is that we would have OWL DL 1.0 entailments  
that would not be OWL DL 1.1 entailments, this means we would have  
breaks in interoperability between our 1.0 legacy and our 1.1.  
proposal in DL - My group makes heavy use of Pellet, but doesn't  
really need most of the new expressivity, so we would find this  
troubling (esp. if we could not upgrade to a new version of Pellet  
because it would break our demos).  In minor cases, carefully  
annotated, this would be ok, but if we do major breaks from previous  
versions this would be a problem.  We see this one as falling  
somewhere in between - because it is hard to tell whether it arises in  
some of the cases we use.  Guess this is that backwards compatibility  
thing -- anyway, guess this is to say in this case we could probably  
live with either solution, but if we do break the old entailments,  
then we would really need careful documentation of where this arises  
in real use (c.f. when we're processing a lot of FOAF stuff with a  
bunch of bnodes through the incremental version of Pellet, should we  
be worried?)
  -JH


On Feb 29, 2008, at 8:54 AM, Michael Schneider wrote:

> Hi Jeremy! Hello David!
>
> Jeremy Carroll wrote on Friday, February 29, 2008:
>
>> One question that came up in HP discussion was whether the proposed
>> resolution to ISSUE-3 would break the monotonicity requirement for
>> semantic extensions
>> http://www.w3.org/TR/rdf-mt/#MonSemExt
>>
>> I suggested that the proposal does break this requirement.
>
> AFAICT, this topic is only relevant for languages which claim to be
> "semantic extentions" as defined by [10]. In this way, RDF semantics  
> is a
> semantic extention to Simple Entailment, RDFS to RDF, and both  
> OWL-1.0-Full
> and pD* are semantic extentions to RDFS. This is the "layered"  
> approach
> which I described in [20]. In fact, the "delta to OWL-1.0-Full  
> idea", which
> we discussed in an earlier telco, would mean that OWL-1.1-Full is  
> becoming a
> semantic extention to OWL-1.0-Full.
>
> But neither OWL-1.0-DL nor OWL-1.1-DL claim to be such a kind of  
> semantic
> extention to RDFS. So you cannot reasonably apply the "General  
> monotonicity
> lemma" from the RDF(S) semantics spec to them, where you have to  
> "Suppose
> that Y indicates a semantic extension of X".
>
>> Here are the
>> test cases, are these correct?
>>
>> Here is a simple test case:
>>
>> A:
>> _:a rdf:type owl:Thing.
>> eg:dp rdf:type owl:DatatypeProperty .
>> _:a eg:dp "foo".
>>
>> B:
>> _:b rdf:type owl:Thing.
>> eg:dp rdf:type owl:DatatypeProperty .
>> _:b eg:dp "foo".
>>
>> A simple-entails B.
>> A simple-entails A.
>
> Yes to both.
>
>> A owl-1.0-dl-entails B.
>> A owl-1.0-dl-entails A.
>
> Not quite sure about this, but I believe to remember Peter saying so  
> in some
> telco. (I also remember that I was surprised about his answer at  
> that time.)
>
>> A does not owl-1.1-dl-entail A.
>> A does not owl-1.1-dl-entail B.
>
> If bNodes will be interpreted as skolems in 1.1-DL, then yes, these  
> will be
> non-entailments.
>
>> If I have understood some of the other aspects of OWL 1.1 DL then  
>> also
>> the following:
>> C:
>> _:c eg:dp "foo".
>>
>> D:
>> _:d eg:dp "foo".
>>
>> C simple-entails C.
>> C simple-entails D.
>> C does not owl-1.1-dl-entail C.
>> C does not owl-1.1-dl-entail D.
>>
>>
>> Note: these show why this change while in keeping with Peter's
>> characterization [1] of the relationship between OWL Full and OWL DL
>> (OWL DL is weaker than OWL Full), misses the relationship
>> between OWL DL
>> and RDFS, which is - on the syntactic subset that is OWL DL, OWL DL  
>> is
>> stronger than RDFS.
>
> This may or may not be the case. Anyway, I would be very reluctant  
> to make
> this "observed" property a requirement for the DL flavours of OWL,  
> now or in
> the future. It has been the basic difference between OWL-1.0-DL and
> OWL-1.0-Full that DL is semantically based on some specific  
> description
> logic (SHOIN(D) in the case of 1.0-DL), while Full is intended to be a
> semantic extention to RDFS. These are two very different design  
> principles,
> so I wouldn't even easily come to the idea that OWL-x.y-DL might be
> semantically a stronger language than RDFS.
>
> Under this pov, even the other direction, that OWL-Full is demanded  
> to be
> semantically a stronger language than OWL-DL, might become  
> questionable. But
> in this case, this property seems at least to be easier to obtain,  
> since
> these two languages share the same OWL specific language features,  
> while
> OWL-Full additionally consists of all the semantic properties  
> inhereted from
> RDFS. So this request looks quite reasonable to me.
>
> Bottom line: I *won't* demand OWL-1.1-DL to be a semantical upper  
> language
> to RDFS, but I *will* ask for OWL-1.1-Full being a semantical upper  
> language
> to OWL-1.1-DL.
>
> Cheers,
> Michael
>
> [10] <http://www.w3.org/TR/2003/WD-rdf-mt-20030123/#DefSemanticExtension 
> >
> [20] <http://lists.w3.org/Archives/Public/public-owl-wg/2008Feb/0068.html 
> >
>
> --
> Dipl.-Inform. Michael Schneider
> FZI Forschungszentrum Informatik Karlsruhe
> Abtl. Information Process Engineering (IPE)
> Tel  : +49-721-9654-726
> Fax  : +49-721-9654-727
> Email: Michael.Schneider@fzi.de
> Web  : http://www.fzi.de/ipe/eng/mitarbeiter.php?id=555
>
> FZI Forschungszentrum Informatik an der Universität Karlsruhe
> Haid-und-Neu-Str. 10-14, D-76131 Karlsruhe
> Tel.: +49-721-9654-0, Fax: +49-721-9654-959
> Stiftung des bürgerlichen Rechts
> Az: 14-0563.1 Regierungspräsidium Karlsruhe
> Vorstand: Rüdiger Dillmann, Michael Flor, Jivka Ovtcharova, Rudi  
> Studer
> Vorsitzender des Kuratoriums: Ministerialdirigent Günther Leßnerkraus

"If we knew what we were doing, it wouldn't be called research, would  
it?." - Albert Einstein

Prof James Hendler				http://www.cs.rpi.edu/~hendler
Tetherless World Constellation Chair
Computer Science Dept
Rensselaer Polytechnic Institute, Troy NY 12180

Received on Friday, 29 February 2008 15:03:10 UTC