W3C home > Mailing lists > Public > www-rdf-interest@w3.org > January 2003

Re: cwm/n3 and naming blank nodes? (calendar rules)

From: Graham Klyne <GK@ninebynine.org>
Date: Fri, 03 Jan 2003 12:24:09 +0000
Message-Id: <5.1.0.14.2.20030103121434.03bb1a70@127.0.0.1>
To: Dan Connolly <connolly@w3.org>
Cc: www-rdf-interest@w3.org, Norman.Walsh@Sun.COM, www-rdf-calendar@w3.org

I think maybe the "problem", if there is one, comes with what you do with 
the results (inferences) from these rules.

Suppose you wish to determine whether a particular date (say, :now) is 
indeed a Birthday corresponding to :E.  Intuitively, if :E has a db:end 
property, and :now is after the value thus specified, then it's not a 
birthday.  But what if there's no db:end property?  I don't think it's safe 
to infer that :now is a birthday unless there is some other explicit 
information that would be inconsistent with a value db:end being before 
:now.  Otherwise, non-mononotic reasoning is invoked.

#g
--

At 12:19 AM 1/3/03 -0600, Dan Connolly wrote:

>Norm Walsh asked, back on 1Dec...
>http://lists.w3.org/Archives/Public/www-rdf-interest/2002Dec/0002.html
>
>about rules for repeating events, some with
>end markers and some without.
>
>I completely missed that message, as I only catch
>up with www-rdf-interest occasionally, but meanwhile, in
>www-rdf-calendar, I wrote some rules that, I think,
>answer Norm's question.
>
>futureEvents.n3: an excercise in processing recurring events
>Dan Connolly (Thu, Dec 19 2002)
>http://lists.w3.org/Archives/Public/www-rdf-calendar/2002Dec/0022.html
>
>Applying the lessons from futureEvents to Norm's birthday
>case, we get:
>
>this log:forAll :p, :s, :o, :t, :u, :l, :k, :m, :E.
>
>{ :p a ab:Contact;
>     p:born :o } log:implies { :p :birthEvent [ a db:Appointment ;
>                                    db:begin-date :o ;
>                                    db:repeat [
>                                       rdf:type db:Repeat ;
>                                       db:frequency "1";
>                                       db:type "Yearly" ] ] } .
>
>{ :p a ab:Contact;
>     p:born :o;
>     p:died :s;
>     :birthEvent :E } log:implies { :E db:end :s }.
>
>
>There's something unsatisfying about this style of rules...
>it feels procedural -- I start to think about
>"creating" a db:Appointment and "modifying" it --
>while writing rules is supposed to be declarative.
>
>I'm not really modifying anything; this is all monotonic.
>But feels wierd.
>
>--
>Dan Connolly, W3C http://www.w3.org/People/Connolly/

-------------------
Graham Klyne
<GK@NineByNine.org>
Received on Friday, 3 January 2003 08:34:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:57 GMT