norman.walsh.name: Comments on /2005/12/16/pimExamplehttp://norman.walsh.name/2005/12/16/pimExample/comments.atom2012-02-13T08:39:10.535677ZComment 1 on /2005/12/16/pimExamplehttp://norman.walsh.name/2010/09/25/oauth#comment00012005-12-17T00:43:01Z2005-12-17T00:43:01ZDan Connollyda39a3ee5e6b4b0d3255bfef95601890afd80709
Some of your rules follow a standardized idiom. For example: 
{ ?p p:livesWith ?q } => { ?q p:livesWith ?p } .
You can just say... 
p:livesWith a owl:SymmetricProperty.
... provided you tell cwm about owl:SymmetricProperty with this rule... 
{ ?prop a owl:SymmetricProperty.
 ?p ?prop ?q } => { ?q ?prop ?p } .
A nice thing about that is that while N3 rules aren't (yet ;-) standardized, OWL is, so now all the OWL tools grok more about p:livesWith, and the p:livesWith a owl:SymmetricProperty. statement only uses the turtle part of N3, which is much more widely supported. e.g. pellet groks turtle and OWL. owl:TransitiveProperty is also standardized. Can't remember if reflexive is standardized... nope; I don't see it in the OWL reference. Likewise, this rule follows an RDFS idiom: 
{ ?p p:spouse ?q } => { ?p p:sigother ?q } .
You can just say... 
p:spouse rdfs:subPropertyOf p:sigother.
and tell cwm what that means thusly: 
{ ?X [ rdfs:subPropertyOf ?PROP] ?Y } => { ?X ?PROP ?Y }.
The semantics of RDFS are completely expressible in N3 rules. swap/util/rdfs-rules is a careful transcription of the RDFS spec into N3. You can't completely capture OWL in N3 rules, but you can express quite a useful bit of it. Jos developed owl-rules.n3, though if you feel all those rules to cwm, it'll probably head into the weeds chasing infinite deduction chains more often than not, so you sorta have to pick and choose. Jos's Euler does better, since it's a backward-chaining reasoner.
Comment 2 on /2005/12/16/pimExamplehttp://norman.walsh.name/2010/09/25/oauth#comment00022005-12-17T13:37:08Z2005-12-17T13:37:08ZDave Pawsonda39a3ee5e6b4b0d3255bfef95601890afd80709

Thanks Norm. This usage does far more (for me) than all the stuff on RDF I've read which, like Uche, has simply had me shrugging.

I'm getting an impression of little piles of classes, namespace lists (extra.n3 etc) and local semantics, some re-used (foaf, dc g), others locally generated (contacts, genealogy). Do they make up what you seem to be calling 'colloquial RDF'? How far is it from what you have to SW RDF? Just a transform?

Shelly's reference to http://www.w3.org/2005/Talks/1214-Trento-IH has made a difference for me. Merging (I guess thats one of your 'wins') is a key item I'm intrigued by.

I'm curious what you are editing with? Emacs I guess, but any major mode for your ownl and N3?

Again. Much appreciated.

DaveP

Comment 3 on /2005/12/16/pimExamplehttp://norman.walsh.name/2010/09/25/oauth#comment00032005-12-17T23:12:05Z2005-12-17T23:12:05ZSusanda39a3ee5e6b4b0d3255bfef95601890afd80709

Maybe it's offtopic, but i just wanted to say, that it's really interesting to read everything this with comments... You discuss here a lot of interesting things on useful themes. Thanks for that.

Comment 4 on /2005/12/16/pimExamplehttp://norman.walsh.name/2010/09/25/oauth#comment00042005-12-20T08:15:57Z2005-12-20T08:15:57ZChristoph Görnda39a3ee5e6b4b0d3255bfef95601890afd80709

Hello Norm, I took your inspiration onto the Mac and reworked some stuff. It's not that complete but it will take your address book on the web.

Comment 5 on /2005/12/16/pimExamplehttp://norman.walsh.name/2010/09/25/oauth#comment00052006-12-07T07:20:46Z2006-12-07T07:20:46ZSteveda39a3ee5e6b4b0d3255bfef95601890afd80709

This is perfect--it's exactly what I have in mind for managing my growing contact list. I notice that you use 'colloquial' RDF; aren't there already URIs for address, contact, etc?