- From: Michael Kifer <kifer@cs.sunysb.edu>
- Date: Wed, 22 Jun 2005 11:15:54 -0400
- To: Jim Hendler <hendler@cs.umd.edu>
- Cc: Jos de Bruijn <jos.debruijn@deri.org>, dreer@fh-furtwangen.de, public-sws-ig@w3.org
Jim Hendler wrote: > > At 12:04 +0200 6/22/05, Jos de Bruijn wrote: > > > >Dreer Michael WI wrote: > > > >>Hello, > >> > >>what is the relationship between WRL and SWRL? > > > >Hello Michael, others, > > > >With WRL we take an approach different from SWRL. We build on the > >experience in > >the Logic Programming and Deductive Database communities (and > >partially other rules communities) and (syntactically) restrict > >ourselves to Horn and add nonmonotonic negation, whereas SWRL stays > >within a First-order framework and simply extends the expressiveness > >of OWL DL. WRL extends a subset of OWL DL; this subset which falls in > >the framework of standard Horn rules. > > > > > > Sorry, but this cannot be right. SWRL assumes closed world semantics > and WRL assumes open world semantics. Thus, we get completely > different entailments in an OWL/RDF world than in a WRL one, so we > are talking something very different than a subset relationship. Ian > Horrocks, Peter Patel-Schneider, and Bijan Parsia (with me as a > kibbitzer) wrote a short paper about this available at [1] **I believe that the mention of OWL-DL in the above message is a typo.** The intent was DLP. Regarding your paper, I wonder what notion of language extension are you using. The one that I was taught as a child says: L1 extends L0 iff L0 \subseteq L1 and for any pair of formulas f1,f2 in L0, f1 |= f2 in L0 if and only if f1 |= f2 in L1. Since neither DLP not RDF has negative queries, the statement about language extension holds for them. Of course, in this case we should use a 1-1 homomorphism to inject L0 into L1, because DLP/RDF have different syntax from WRL. > Let me be clear, I'm all for Web (and Sem Web) rules languages, but > if it isn't open-world, I don't see how it can be Sem Web, since it > violates the base assumption on which all of RDF, RDFS, and OWL sit. > This is easily fixable, and at the Rules workshop the idea of a > "Scoped Negation as Failure" was developed to handle this -- I'd love > to see WRL (and SWRL) extended to have a SNAF mechanism, because then > we don't violate the basic principles of the Web architecture and the > Semantic Web, but we should be precise - two things with very > different Semantics and entailments cannot be referred to as subsets > of each other. I would like to mention here that the notion of scoped negation was *NOT* developed at that workshop. It existed before, and I don't know who first coined it (must have been 10 years or so). This type of negation is even implemented in some systems, like FLORA-2. In that particular system it exists since the end of 2000. --michael > -JH > > > > > [1] http://www.cs.man.ac.uk/~horrocks/Publications/download/2005/HPPH05.pdf > > >Best, Jos > > > >> > >>Thanks in advance > >> > >>Michael Dreer > >> > >> > > > >> Steve Ross-Talbot wrote: > >> > >> >>Out of curiosity and also business pragmatics what is the relationship > >> >>between WRL and RuleML? > >> > >> > >> > >> RuleML provides an extensive lattice of sublanguages ranging from > >> production rules to regular LP rules to First-Order Logic, es well as > >> sublanguages with such features as slotted syntax, meta-programming, and > >> courteous logic programming. > >> > >> WRL makes a choice to use include particular features in the language. > >> It turns out that these features correspond to certain RuleML > >> sublanguages to a great extent. Therefore, RuleML can be used for the > >> XML serialization of a large part of the language. > >> > >> The following document describes the (RuleML and other) schemas used for > >> the XML serialization of WRL: > >> > >> http://www.wsmo.org/wsml/wrl/wrl-xmlschemas.html > >> > >> More details on the XML serialization of WRL can be found at: > >> > >> http://www.wsmo.org/wsml/wrl/wrl.html#cha:wrl-xml > >> > >> I hope this clarifies the issue. > >> > >> > >> Best, Jos > >> > >> > >> >>Cheers > >> >> > >> >>Steve T > >> >> > >> >>On 21 Jun 2005, at 08:28, Jos de Bruijn wrote: > >> >> > >> >>Dear all, > >> >> > >> >>On behalf of WSML, AIFB and NRCC I am pleased to announce: > >> >> > >> >>The Web Rule Language WRL - version 1.0: > >> >> > >> >>http://www.wsmo.org/wsml/wrl/ > >> >> > >> >> > >> >>WRL is a rule language for the Web and the Semantic Web and was > >> >>heavily influenced by the Web Service Modeling Language WSML > >> >>(http://www.wsmo.org/wsml/wsml-syntax). > >> >> > >> >> > >> >>Best regards, > >> >> > >> >>Jos de Bruijn > >> >> > >> >>>> > >> >>>> > >> > >> > >> -- > >> Jos de Bruijn, http://www.uibk.ac.at/~c703239/ > >> +43 512 507 6475 jos.debruijn@deri.org > >> > >> DERI http://www.deri.org/ > >> ---------------------------------------------- > >> > >> Only two things are infinite, the universe and > >> human stupidity, and I'm not sure about the > >> former. > >> -- Albert Einstein > > > > > >- -- > >Jos de Bruijn, http://www.uibk.ac.at/~c703239/ > >+43 512 507 6475 jos.debruijn@deri.org > > > >DERI http://www.deri.org/ > >- ---------------------------------------------- > > > >Only two things are infinite, the universe and > >human stupidity, and I'm not sure about the > >former. > > -- Albert Einstein > >-----BEGIN PGP SIGNATURE----- > >Version: GnuPG v1.2.5 (GNU/Linux) > >Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org > > > >iD8DBQFCuTeV4lqeiwiiHN4RAm8yAJ491RyZZhgDHODjtbIiwT/esmS3IgCfZ3Ws > >xyLcZkDOCDxFNHoHi/DLVeQ= > >=UHcZ > >-----END PGP SIGNATURE----- > > -- > Professor James Hendler Director > Joint Institute for Knowledge Discovery 301-405-2696 > UMIACS, Univ of Maryland 301-314-9734 (Fax) > College Park, MD 20742 http://www.cs.umd.edu/users/hendler > >
Received on Wednesday, 22 June 2005 15:16:14 UTC