Re: Input sought on datatyping tradeoff


You are making life a little difficult for me here :)

You reject the constraints on the answers put in the question, which is 
fair enough and we should discuss your reasons for that.  What I'm looking 
for though is a judgment about which of the two 'features' is more 
important to have.  If it later turns out we can have both, then fine, but 
if we had to choose between A being yes and D being yes, which do we go for?

I'm currently interpreting your answer as favouring what we call the 
'untidy' position.

In this position two literal nodes in an RDF graph with the same label, 
e.g. "10" need not denote denote the same thing in the domain.

You also suggest that this position can be reconciled with test case A1 
being answered YES.

I must ask you, if it were not possible that the answer to A1 could be YES, 
would you still hold the same opinion.  I've tried to demonstrate a problem 
with your current position in:

I add some further comments to your message below.


At 18:12 12/07/2002 -0400, Drew McDermott wrote:

>I have posted on rdf-logic why I don't think the questions below are
>the right questions.  However, I can answer them under the assumption
>that they are the right questions.
>    A few simple test cases:
>    Test A:
>       <Jenny> <ageInYears> "10" .
>       <John>  <ageInYears> "10" .
>    Should an RDF processor conclude that the value of the ageInYears
>    properties for Jenny and John are the same?
>Yes, because no matter how the argument to <ageInYears> is parsed, the
>same input string must yield the same literal.

We appear to have different assumptions about what is going on here.  I 
don't know enough to grok what your viewpoint is so I'll express this as 
some observations that jar with me.

"Parsed"?  We are talking about model theory here.  There is no parsing.

>   (Escape clause: If
><ageInYears> is allowed to consider its first argument,

"argument"?  There are no arguments.  <ageInYears> is a 
property.  "consider" is not something it does?  <ageInYears> is part of an 
expression in a language.

>  then I switch
>to "No conclusion."  Perhaps <Jenny> and <John> are entities of
>different types, which parse


>their <ageInYears> properties
>differently.  I assume this is not allowed.)
>    There are variations on this test which should be considered before 
> answering.
>    Test A2:
>       <Jenny> <ageInYears> "10" .
>       <Jenny> <testScore>  "10" .
>    Should an RDF processor conclude that the value of Jenny's ageInYears
>    property is the same as the value of Jenny's testScore property?
>No.  Nor can it conclude that they are different.
>    Test A3:
>       <Jenny> <ageInYears>   "10" .
>       <Film>  <title>        "10" .
>    Should an RDF processor conclude that the value of Jenny's age 
> property is
>    the same as the value of the Film's title property?
>No.  As in case A2, it can't conclude they're not the same either.
>    The answer must be the same for all three of these A tests.
>I disagree; see above.
>    These test cases only relates to the situation where there are no range
>    constraints on the properties.
>Assuming nonmonotonicity is to be avoided, that's why we have to
>respond "I don't know" in cases A2 and A3.  The situation in which
>"there are no range constraints" is undefined in RDF, because you
>could always discover some range constraints later, and nothing
>concluded earlier can be removed.

I'll show something pretty close to non-monotonicity in for A1 as you have 
it in another thread in RDF Logic.


>    Now for a different kind of test.  How do the values of the two idioms 
> relate?
>    Test D:
>       <Jenny>      <ageInYears> "10" .
>       <ageInYears> rdfs:range xsd:decimal .
>       <John>  <ageInYears>   _:a .
>       _:a     xsdr:decimal   "10" .
>    Should an RDF processor conclude that Jenny and John have the same
>    age?
>    [Note: in this example the range constraint is expressed using
>    rdfs:range.  We may have to introduce a special datatyping range 
> property,
>    but that is an independent detail for now.]
>    It is not possible to have the answers to Tests A and Test D both be
>    yes.  Either the A's can be yes or D can be yes, but not both.  We 
> have to
>    decide which of these is the most important to have.
>I don't see why they can't both be Yes.  (I've read Brian's message of
>July 1 arguing that they can't, but I'll respond on the rdf-logic
>mailing list.)
>                                              -- Drew McDermott

Received on Monday, 15 July 2002 14:29:34 UTC