Re: Web Semantics of Datasets (v0.2)

On Mon, 2011-10-10 at 15:19 -0500, Pat Hayes wrote:
> On Oct 10, 2011, at 6:30 AM, Sandro Hawke wrote:
> 
> > Here's some revised wording for the proposal, getting a bit closer to
> > spec text.   It's still somewhat informal, and mixing normative and
> > non-normative bits, and best-practice.   And it's not as clear as it
> > should be about handling change over time.
> > 
> >    -- Sandro
> > ===
> >  A dataset D is true
> 
> You have to say, true in an interpretation, and modify the rest of the phrases accordingly.
> 
> > iff (1) its default graph is true and (2) for
> >  every pair of <N,G> in D, N names something (a "resource", sometimes
> >  called a "g-box") which, at every time T in R, has G as its current
> >  state.
> 
> What is R? 

R is a set of points in time, as discussed elsewhere.   To me that seems
like a perfectly reasonable way to think about this, but *shrug* feel
free to come up with some other way to keep graphs in sync.  (see
example at the end of this email.)

> > 
> >  It follows from AWWW that if N is an IRI which can be dereferenced,
> >  a successful, correct dereference of N at any time T in R must yield
> >  a serialization ("representation") of G.
> 
> No, that does not follow. There is no immediate connection between being a name of something and dereferencing to something. If we want this relationship, we have to state it explicitly. Also, you have stated this using a modality ("can be") which opens up a huge can of worms. (What if it could be dereferenced but never, in fact, was? Did it name the graph under those circumstances?) I would strongly suggest that we avoid this kind of language.

The heart of ReST and AWWW is that dereferencing an IRI yields a
representation (serialization) of the state of the thing named by that
IRI.   That's the "immediate connection" I'm using.  So, doesn't it
follow from that?    

> >  In order to know whether a dereference occurs at a time in R, it is
> >  useful to have R declared in the default graph of D, or in another
> >  nearby, easy-to-find data source.  Where possible, is is helpful to
> >  have R be All Time; that is, having N name a resource whose state,
> >  by definition, never changes.
> 
> Why is that the same condition? Surely this begs the question of whether or not G can change. (?)
> 
> > 
> >  In RDF data, N may be used (1) directly, to name the g-box,
> >  expressing things like the license that applies to its state, or who
> >  controls it; and (2) indirectly, to refer to G as the current state
> >  of the g-box.  
> 
> How are these two uses distinguished? There is currently no provision for this kind of indirection in the RDF semantics. 
> 
> > Indirect reference can be used to express things
> >  about an RDF Graph (a "g-snap"), like that it was the graph some
> >  entity asserted at some time.  Indirection is done in the semantics
> >  of the predicates with which N is used.
> 
> ?? It is? How? 

Sorry, I didn't mean anything complicated, and I don't think it needs
any changes to the RDF Semantics.

An example of direct use would be:

  <http://www.w3.org/People/Berners-Lee/card> eg:author
<http://www.w3.org/People/Berners-Lee/card#i>

because there I'm talking about the g-box that is Tim's maintained
"card".

An example of indirect use would be:

  <http://www.w3.org/People/Berners-Lee/card> eg:boxtriples 223.

by which I mean the graph (now) in that g-box has 223 triples.  That's
"indirect" because I'm talking about the graph in the g-box, not the
g-box.  

Perhaps that should be formalized as two triples:
 

  <http://www.w3.org/People/Berners-Lee/card> eg:contents _:g.
  _:g eg:triples 223.

That's probably clearer, but harder to use.   boxtriples can be seen as
just the property-chain of (contents,triples).

> >  When N is used indirectly, the reference to G only holds inside time
> >  range R, of course.  Care must be taken not to use N as if it
> >  necessarily referred to G, outside of R.  Since R is defined to be
> >  the same for all elements of D, indirect reference is safe in the
> >  default graph.   
> 
> How does time-dependent reference work in the semantics? (Or is that my job to figure out? If so, I can tell you now it will be a major change, which will require rewriting the OWL and RIF standards, among others.)

I'm not asking for anything different in the RDF semantics than is in
there now with the way I can currently say in RDF:

  <http://www.w3.org/People/Berners-Lee/card> foaf:age 56.

That works fine in the RDF semantics today, doesn't it?   All I'm asking
for is a bit more understand of how it works, and some vocabulary for
machines to share the metadata they need to keep things straight when
they fetch that kind of thing from different places at different times.
(And then I also want to apply it to dereferencing.)

   -- Sandro

Received on Tuesday, 11 October 2011 01:54:34 UTC