- From: pat hayes <phayes@ihmc.us>
- Date: Thu, 15 Jan 2004 17:48:11 -0600
- To: "Jos De_Roo" <jos.deroo@agfa.com>
- Cc: www-archive@w3.org
>[Pat, I'm cc'ing www-archive as it's easier that way to point back] > OK <snip> > > Was there a particular question you had that I >> can focus on? > >Well, the backbround was an asking of your opinion >about the last message of last year in www-archive >http://lists.w3.org/Archives/Public/www-archive/2003Dec/0047.html >Is there an aternative way of *proving* inconsistencies ?? >(our usage of log:inconsistentWith is what >euler uses for the OWL test cases and also >in some observer component that we test) You asked there if this misuses anything. I don't think so. All of these should logically be equivalent, but the exigencies of search might make some more effective than others: A inconsistentWith B (A and B) entails (false) where 'false' could be any explicit contradiction, in principle, but of course you would need a strongly complete system to be sure of getting every possible contradiction actually derivable, so should likely choose a single representative impossiblity-indicator such as (P and not P). (A and B) entails not(A) C: (A consistentWith B) not C Not sure if this is any help.... Pat Oh, PS: proving inconsistencies. The only real technique is to derive an explicit contradiction, like the null clause in resolution. The 'xml clash' thingie in RDFS was an example closer to home. You have to figure out the ways that inconsistencies can arise, in practice, and kind of nudge your search engine in those directions. -- --------------------------------------------------------------------- IHMC (850)434 8903 or (650)494 3973 home 40 South Alcaniz St. (850)202 4416 office Pensacola (850)202 4440 fax FL 32501 (850)291 0667 cell phayes@ihmc.us http://www.ihmc.us/users/phayes
Received on Thursday, 15 January 2004 18:48:13 UTC