W3C home > Mailing lists > Public > public-owl-wg@w3.org > February 2008

Re: Fragments discussion, continued

From: Jim Hendler <hendler@cs.rpi.edu>
Date: Thu, 21 Feb 2008 10:07:08 -0500
Cc: "'Web Ontology Language \(\(OWL\)\) Working Group WG'" <public-owl-wg@w3.org>
Message-Id: <D037B954-F5E9-4A6F-BB54-0C12C7EEA9C9@cs.rpi.edu>
To: Boris Motik <boris.motik@comlab.ox.ac.uk>
Apologies that I missed the meeting, unavoidably detained
  I think we're making good progress - one of the things I'd push for  
is to try to see if we can unify things down to a smaller number of  
entities (possibly with Full/Dl realizations) - so as I see it now
  1 - We have a plan for OWL Lite - which is to maintain as is to the  
extent possible
  2 - EL++ is a fragment of DL - I believe the argument is that it  
scales better for Tbox than OWL DL - is that right?  (If not, I  
apologize -There was email from Carsten saying a lot of people were  
using it, but it didn't say explicitly while it was different from DL  
in terms of formal properties - however the fragment document makes it  
clear it is PTIME-complete for the Tbox operations (it is "not  
applicable" for Abox).   One thing that will help is making the  
argument for using this over DL per se (i.e. when to choose which) -  
we should make sure we address this in the fragments document.  (to be  
clear, this is not meant to be anti-EL or anti-DL -- it's that we need  
to help users understand which one they might consider using given  
various conditions, otherwise we end up with EL++ and DL competing, as  
it were, for the same market share)
  3 - OWL prime  (suitably renamed - I still sort of like RDFS 3.0) is  
argued for by adopter need and the need for interoperability in the  
"web app development" space.  Someone asked at the cal if there were  
multiple interoperable  implementations and the answer is yes,  
definitely. Also scalability, and as someone mentioned on the IRC,  
it's been used for Aboxes of 10^9 triples, so I think we can say yes.
  4 - Alan argues for a well defined fragment that would have provable  
abox scalability - I assume he means provably so against some measure  
(like completeness) ala datalog reduction or such.

Personally, I'm not yet convinced on this 4th one.  If we use  
Carsten's 3 crieria for a fragment, we ask for:
  1 - supported by a group of users
  2 - favorable property that separates them from Full and DL
  3 - maximal w/respect to that property

Alan, can you explain who it is whose been asking for that fragment  
(as separate from the Prime fragment) and what the distinguishing  
property would be?



>> -----Original Message-----
>> From: public-owl-wg-request@w3.org [mailto:public-owl-wg-request@w3.org 
>> ] On Behalf Of Alan Ruttenberg
>> Sent: 19 February 2008 15:27
>> To: Web Ontology Language ((OWL)) Working Group WG
>> Subject: Fragments discussion, continued
>> It seems (I hope) we are getting closer to some sort of mutual
>> understanding of our needs for fragments. I'd like to sketch out the
>> shape of what I'd be happy to see as an outcome of tomorrow's  
>> meeting.

Received on Thursday, 21 February 2008 15:07:42 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:42:02 UTC