[sc34wg3] And yet another...

[Lars Marius Garshol]

* Holger Rath
| 
| Well, Graham introduced me to quadruples 1.5 years ago when he
| developed the Semantic Web Server. So the idea was familiar to me
| :-)

Oh, you're a bad test subject, then. :)

I knew Graham was onto this, too. We discussed this at least once when
he worked for us (after he worked on the SWS), and I think I came up
with this myself (though to be frank I am not absolutely sure), and
that when presenting it to him I realized from his reaction that he'd
already worked out pretty much the same thing for himself. I was not
aware that he'd implemented it, though.
 
| Yep. And you sketched already how to do it: define a vocabulary for
| the TMDM properties. Maybe using RDFS and/or OWL.

Something along those lines, yes. The RTM vocabulary for mapping RDF
to topic maps is one possibility. What you suggest (and which I admit
was in the back of my mind) is another. I don't think either is quite
*there* yet, but I do think it can be done.

| [TMRM relationship]
| 
| That's what I meant. And the TMRM assertions always reminded me of
| RDF statements. And with the 4th element in the tuple you build the
| bridge for reification.
 
Right. I suppose the main difference is that quad statements are
fixed-size, and that theirs are variable-size (since you can always
throw in more roles and players). However, given that I represent each
role assignment as a quad it could be argued that my associations are
variable-sized, too. The difference is that I distinguish between
occurrence/name assignments and associations, which TMRM and the Tau
model do not do. Whether that's a strength or a weakness I'm not sure.

| I am wondering how the TMRM people will react to your proposal.

So am I. :-)

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >