W3C home > Mailing lists > Public > public-owl-dev@w3.org > July to September 2011

Re: A solution to integrate CWA into OWA.

From: Pat Hayes <phayes@ihmc.us>
Date: Fri, 19 Aug 2011 15:12:59 -0500
Cc: Parisa <bparsia@cs.man.ac.uk>, <public-owl-dev@w3.org>
Message-Id: <E8D82A18-C46D-4718-BF09-87ED4C34A80C@ihmc.us>
To: duanyucong <duanyucong@hotmail.com>
Permit me to try to help with some explanations. 

Yucong, greetings. Bijan is correct in that it is rather hard for us to see what exactly it is that you are saying, and that you seem to be using familiar technical words in ways that do not correspond to their accepted meaning. As I share Bijan's state of confusion regarding your exact meaning, you will have to bear with me if I misunderstand you. However, I will try to respond to your questions as best I can. 

Your thread began by contrasting the CWA and the OWA, and trying to suggest some way in which they could be resolved or unified. However, your statements in these early emails strongly suggest that your understanding of what these terms (OWA and CWA) mean, does not correspond to their actual meaning in the technical literature which uses them. In particular, it is quite wrong to say, as you seem to be saying, that the OWA is somehow incompatible with negation or requires a mental stance which denies the validity of negation. There are logics which either change the meaning of negation (intuitionistic logic, for example) or which refuse to use negation at all (positive logics) but none of these are directly related to ontology formalisms such as OWL or indeed to the whole semantic-web field. 

The closed world assumption is the assumption that the entities which are named are all the entities that exist, and all the ground facts that are true of them are actually known. Put another way, it is the assumption that everything has a known name, and all ground facts are known. (By 'ground' I mean simple assertions without any variables or quantifiers in them, the things one might find in a data base table.) This has the consequence that if you search through all the facts and fail to find something, then you know it must be false: a rule often called negation-by-failure. This is a good way to reason when you have some confidence that the list of  facts you are working with is complete (in some sense), for example if one has a database of company employees: to fail to find a name in the list is a good indicator they are not an employee.

Open world reasoning is simply any reasoning which does not make the closed-world assumptions. For example, if I tell you that Bill and Harry work for Raytheon, you should probably not conclude that someone else, say called Jim, does NOT work for Raytheon simply because he is not listed in the list of names I happen to mention. Similarly you should not conclude that Raytheon has only two employees. 
Nevertheless, you can of course (as Bijan has amply demonstrated) correctly reason with negation: the only thing you should not do is *infer* a negation from a mere failure to find some piece of data. It is often summarised by a slogan like, you might always learn something new in an open world. Just because you don't know that X, you cannot ( in an open world) conclude not-X; or at any rate, if you do draw such a conclusion, you are running a risk of stating a falsehood.  

Does this help? As I hope you can see, this OWA/CWA distinction does not have the rather far-reaching metaphysical dimensions that you seem to think it has. 

Pat Hayes

On Aug 19, 2011, at 11:12 AM, duanyucong wrote:

> Dear Bijan,
>  
> I have repeated that i am learning OWL.
>  
> But i can not understand that "Regardless of whether I'm an authority, I am correct ...".
> Do you mean that your subjective judgement will be reliable than others' ? How ?Based on what? your subjective judgement...
>  
> also from your "I cannot determine from your emails a cognitive structure that is 1) coherent and 2) reasonably in accordance with standard understanding ... "
> here i have to repeat again, i am a learner of OWL if it is the source do the standard understanding instead of the subjective but judical understanding in your mind.
>  
> Sincerely,
>  
> Yucong
> 
>  
> From: bparsia@cs.man.ac.uk
> Date: Fri, 19 Aug 2011 16:33:10 +0100
> To: duanyucong@hotmail.com; public-owl-dev@w3.org
> Subject: Re: A solution to integrate CWA into OWA.
> 
> On 19 Aug 2011, at 15:10, duanyucong wrote:
> 
> Dear Bijan,
>  
> You know that i will agree with you when you say what have expressed is not conform to semantics of OWL specifications.
> 
> I do?
> 
> I would assume that when you said so, you are an authority.
> 
> Regardless of whether I'm an authority, I am correct in this instance. Yay me!
> 
> However i can not understand whether so many judgement expressed by you to a student
> 
> My mind reading capabilities failed to detect that you are a student. But I don't see how that's relevant: Students, in my experience, are quite as capable of writing gibberish as professors.
> 
> such as "Gibberish", "silly",...etc are still based on 
> specification of OWL specification? Or some other AI specifications?
> 
> It's based on the fact that I cannot determine from your emails a cognitive structure that is 1) coherent and 2) reasonably in accordance with standard understanding of the terms you used. Part of this is your trouble with English which, I presume, lead you to state things a bit less tentatively than you might wish, But part of it is clearly a conceptual problem. AFAICT, no concept you mobilized was used in any standard or reasonable sense. (E.g., "notation", "CWA", "OWA", "semantics", "ontological", "negation"). This is characteristic of naive stud ents and of kooks.
> 
> BTW: I am learning OWL. So please expect that i will make mistakes and have questions:-)
> 
> No worries! I do suggest that you ask questions rather than making pronouncements. You might also recognize that you are deep in the weeds so it might be a good idea to draw back, read a textbook (or the OWL specs) with a fresh eye, and try again. If you can find a local expert to chat with face to face you might clear up you conceptual confusions more quickly. But it could help to start by not dismissing standard presentations of the CWA.
> 
> Cheers,
> Bijan "With your best interests at heart" Parsia.

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
Received on Friday, 19 August 2011 20:13:49 GMT

This archive was generated by hypermail 2.3.1 : Wednesday, 27 March 2013 09:32:59 GMT