W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > July 2001

Re: A use case for anon nodes - action from telecon

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: Thu, 26 Jul 2001 23:34:46 +0100
Message-ID: <3B609B06.D2571FF@hplb.hpl.hp.com>
To: pat hayes <phayes@ai.uwf.edu>
CC: w3c-rdfcore-wg@w3.org


pat hayes wrote:

[...]

> No, of course not. But the original example seemed to involve some
> kind of transaction or handshaking (?) going on between the buyer and
> the seller, mediated by some kind of relationship between the RDF
> smears (good word!) they have published, ie this RDF-publishing is
> supposed to be in some commercial context of actual activities. 

It seems as though the context has some significance in how we
view this scenario.  The participants have intent to establish
commercial transactions.  Lets try a thought experiment, change
the scenario and see how that affects how we look on it.

Two folks, Anna and Bertie have home pages on the web.  For no
other reason than its the done thing to do to be hip, they
include in their home pages RDF statements about where they live
and where they work e.g.

  <Anna>      <livesAt> _:AnnasHome .
  _:AnnasHome <zipcode> "12345" .
  <Anna>      <worksAt> _:AnnasWork .
  _:AnnasWork <zipcode> "54321" .

  <Bertie>      <livesAt> _:BertiesHome .
  _:BeritesHome <zipcode> "12345" .
  <Bertie>      <worksAt> _:BertiesWork .
  _:BertiesWork <zipcode> "54321" .

Pure assertions, yes?  No query, yes?  No intent to create 
a relationship, yes?

Friends of the Planet have noticed this hip behaviour and they
have a web crawler that sucks up this information and looks
for folks with close home and work zip codes and sends them
email suggesting they might like to car share.

Seems to me that the processing that goes in this scenario and
the original are effectively the same.  The point is, that the
query, if there is one, is a feature of a third party, not of
the publishers of the RDF.

BTW, Anna has two homes, both with the same zipcode.


> Once
> we talk about things that are done with, or as a result of,
> publishing some RDF, we need to be clearer about what relationships
> are being assumed between the RDF that is being published. There is a
> real difference between asserting an existential (saying something
> exists) and holding up an existential as a challenge for someone else
> to prove (eg saying I need someone to sell me some roses, and
> expecting this to somehow extract all the RDF that talks about
> selling roses.) (Eg skolemisation is OK in the former case, not the
> latter; variables can get bound at inference-time in the latter case,
> not the former.) If both the buyers and the sellers are just making
> assertions, then they can both draw all sorts of conclusions, but
> nobody is going to discover anything about themselves that they didnt
> already know, so no roses are going to actually get sold.

I don't dispute that different processing may be required in
different circumstances.  But those circumstances may not depend
on the intentions of the publishers of the RDF.

Does this shed any light on where I'm going wrong?
My brain is beginning to hurt.

Brian
Received on Thursday, 26 July 2001 18:37:28 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:38:18 EDT