Re: Merging Databases

On Jul 21, 2009, at 8:42 PM, Peter Ansell wrote:

> 2009/7/22 Pat Hayes <phayes@ihmc.us>:
>>
>> On Jul 21, 2009, at 7:26 PM, Alan Ruttenberg wrote:
>>
>>> On Tue, Jul 21, 2009 at 1:23 PM, Toby Inkster<tai@g5n.co.uk> wrote:
>>>>
>>>> On Tue, 2009-07-21 at 19:52 +0300, Bernhard Schandl wrote:
>>>>
>>>>>> I would say: Never assert sameAs. It's just too big a hammer.
>>>>>> Instead use a wider palette of relationships to connect entities
>>>>>> to other ones.
>>>>>
>>>>> which ones would you recommend?
>>>>
>>>> skos:exactMatch = asserts that the two resources represent the same
>>>> concept
>>
>> Say, refer to the same thing.
>>
>>>> , but does not assert that all triples containing the first
>>>> resource are necessarily true when the second resource is  
>>>> substituted
>>>> in.
>>>
>>> I'm having trouble parsing this one. I don't know what concepts are,
>>> but they are an odd sort of thing if they can be the same, but can't
>>> be substituted.
>>
>> This is exactly what is needed in many cases. Philosophical  
>> terminology is
>> that they have the same referent but not the same sense, and lack of
>> substitutability reflects the unfortunate but inevitable fact that  
>> the Web
>> as a whole is not referentially transparent (yet). More mundane  
>> example, the
>> same person might need to be referred to in one way in one context  
>> and
>> differently in another, just because the two social contexts require
>> different forms of address. (That example from Lynn Stein.)
>
> The two things may also still be described using the same sense but
> the representations could be structurally incompatible and the
> substitution has risky effects even though one is sure that the two
> representations mean the same thing in the same sense.

I was using 'sense' in Frege's, er, sense (in the original, Sinn vs.  
Bedeutung), so I don't believe that what you say can be correct. If  
two expressions have the same **sense** then they really are  
intersubstitutable in all contexts, I believe. But maybe you have a  
counterexample? (Lynn's 'social use' case might be one, in fact.)

>
> To me it is partly an issue of data granularity, although it isn't
> necessarily heirarchical, so can't be generally represented and
> resolved using rules. Something like
> myterms:alternateRepresentationUri might be what I would see it as, in
> addition to the myterms:alternateSenseUri that you described. (The
> could both be sub properties of rdfs:seeAlso and/or something like
> myterms:alternateUri, without any harm as far as I can tell) The term
> rdfs:seeAlso provides this in some way but going one step further
> without implying any extra information inside of the system would be
> nice as seeAlso has been used to point people to web page addresses
> that wouldn't actually be substituted by people even if they wanted to
> so its history is too broad for the purpose here.

Yes, seeAlso is really extremely weak and can mean anything.
>
> Ideally one would never have these issues because people can now
> communicate unambiguously and in realtime, but there are still people
> that are needed to put the rules in and encode the information. They
> may quite easily fundamentally disagree on a single representation and
> its broader implications so there will be cases where you want to say
> real world equivalence without implying any extra information that
> will disturb the system in undesired ways. If people want to describe
> without automatic implication they should be able to in my opinion.

Well, of course they can, its a free Web. But what does this mean? A  
description amounts to some kind of constraint, thought of  
semantically. When you describe the way things are, you rule out other  
ways they cannot be. And then that ruling out automatically supports  
certain kinds of entailment. Even if you decide not to draw those  
conclusions, they are still **valid**, which is all that the specs  
specify, in fact. So you can't describe without automatically  
entailing, even if you don't do anything operational with those  
entailments. You cannot describe without implying some other things,,  
by the very nature of describing.

> If
> you have the datasets locally stored you could always locally redefine
> the term to be a sub property of owl:sameAs if you needed to see what
> would happen if complete automatic implications were made based on
> these statements.

True, and in fact you could do this many other ways, purely locally. I  
agree that local experiments are completely unconstrained by the Web  
specifications, which really only apply to **published** content.

Pat

>
> Cheers,
>
> Peter
>
>

------------------------------------------------------------
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, 24 July 2009 16:55:27 UTC