Re: RDFCore WG: Datatyping documents

On 2002-02-04 21:41, "ext Damian Steer" <D.M.Steer@lse.ac.uk> wrote:


> Actually I was confused. Jeremy's MT was doing what I suggested. It is
> the pairing itself that confuses me.
> 
> So rather than denoting < lexical value, value > you want <datatype,
> lexical value>? I can't see that that helps at all. < float, "0" > !=
> < float, "0.0" >.

TDL pairings that denote the same value are not garunteed to be equal.

It must be kept in mind that RDF has no built-in native datatypes
and hence no internalized canonical representation for values.

All that one can do is to specify a lexical form along with
the datatype to which it corresponds, and together, these
denote one and only one value.

The only way you can be sure that two TDL pairings that denote
the same value are in fact the same pairing is to use canonical
lexical forms. But there is no guaruntee that a lexical form
will be canonical, thus we must always presume an N:1 mapping
from lexical to value space.

You might say that there are two levels to the TDL proposal,
the TDL pairing of lexical form and datatype <"0", float> and
the mapping pair of lexical form and value <"0", 0>. The former
resides in the graph. The latter in the MT interpretation of
the graph. A TDL pairing corresponds to one and only one mapping
and the mapping denotes the value.

> But perhaps you mean <float, "0"> is syntactic, to be interpreted
> as the float 'function' (denoted by 'float') operating on "0", i.e. 0?
> Which sounds right, if you can fiddle it for the case of
> non-functions.

This, I think, is a good way of thinking about the basic concept
of TDL pairings (not the mapping pairs in the TDL MT).

The datatype serves as a context (or function) by which the literal
(lexical form) recieves an interpretation and from which we get
a value.

Rather than writing the TDL pairing <"lit", datatype> = value,
you could also express it as datatype("lit") -> value

Cheers,

Patrick

--
               
Patrick Stickler              Phone: +358 50 483 9453
Senior Research Scientist     Fax:   +358 7180 35409
Nokia Research Center         Email: patrick.stickler@nokia.com

Received on Monday, 4 February 2002 15:46:50 UTC