W3C home > Mailing lists > Public > www-rdf-calendar@w3.org > December 2002

RE: RDF semantics, access control description and timeranges

From: Geoff Chappell <geoff@sover.net>
Date: Mon, 16 Dec 2002 11:27:53 -0500
To: "'Dan Connolly'" <connolly@w3.org>
Cc: "'www-rdf-calendar'" <www-rdf-calendar@w3.org>
Message-ID: <00bc01c2a520$18b11470$835ec6d1@GSCLAPTOP>



> -----Original Message-----
> From: www-rdf-calendar-request@w3.org [mailto:www-rdf-calendar-
> request@w3.org] On Behalf Of Dan Connolly
> Sent: Monday, December 16, 2002 10:40 AM
> To: Graham Klyne
> Cc: www-rdf-calendar
> Subject: Re: RDF semantics, access control description and timeranges
> 
> 
> I hope to reply in substance presently, but
> I'm not sure I'll be able to...
> 
> On Sun, 2002-12-15 at 07:33, Graham Klyne wrote:
> [...]
> > It seems that, using RDF, it is difficult to construct usage
scenarios
> in
> > which adding additional properties can be used to refine the
precision
> of
> > what is specified.  To do that requires a form of default reasoning,
> which
> > is non-monotonic.
> >
> 
> Indeed, this is a recurring issue; I'm working on proving
> that my new calendar knows everything my old calendar knew,
> and I'm wrestling with this issue; in particular, I have
> rules ala
> 
> 	IF it's an event with a date-time and timezone, do x
> 	ELSE do Y
> 
> well, I can't tell that it's *not* an event with  a date-time
> based on lack of information.

It seems reasonable that the real-world apps on the edges of the
semantic web (i.e. the things that are purely sinks and not sources)
will inevitably have to sometimes make closed world assumptions (even if
it's just to normalize imported data from a particular source). After
all, there's nothing wrong with semantics in application code at some
point :)

We have a switch/case construct in rfdql that we put in after battling
extensively with this issue. An example use:

Infer {[ex:title] ?s ?o} from 
	{[rdf:type] ?s [ex:something}
	and switch(?s)(
		case {[rdfs:label] ?s ?l}:
			?o=?l
		case {[rdfs:comment] ?s ?c}
			?o=?c
		default:
			?o=?s
		);

which will only ever return a single value for ex:title per subject.

It's definitely non-monotonic in the global sense but at least is
monotonic from the perspective of the query (i.e. if a later rule fires
during query execution that changes results, the previous intermediate
results and all of its sole dependants will be retracted).

--geoff chappell


> 
> I'm using log:notIncludes to do it.
> 
> More later, I hope.
> 
> 
> > I'm left thinking that some practical guidelines are required to
avoid
> such
> > potential problems in general RDF use.
> >
> --
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Monday, 16 December 2002 11:28:20 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:14:11 UTC