[rdfweb-dev] seeAlso's for FoaF documents

Dan Brickley danbri at w3.org
Sat Mar 1 12:34:47 UTC 2003


* Morten Frederiksen <mof-rdf at mfd-consult.dk> [2003-03-01 12:40+0100]
> On Saturday 01 March 2003 11:47, Danny Ayers wrote:
> > hasFOAFDocument subProperty seeAlso
> I agree with the principle, and since my own scutter is only on the drawing 
> board, including that kind of inference, I wouldn't mind it at all.
> 
> However, it does leave the less advanced scutters (currently: all?) behind, 
> so I'm not sure that's the best way to do it (yet).
> 
> "Typing" the object of the seeAlso seems to be the best way to do it at this 
> point in time, the less advanced scutters would simply ignore it.

Yes, I think we should be careful to leave the basic hypertext infrastructure
stuff so that it works without consumers of the doc needing to (a) 
dereference schemas (b) do subPropertyOf reasoning. So yes, decorating the 
thing pointed to with additioanl properties seems a better design. It all
makes for greater expressivity: we can add several different properties,
types etc depending on scenario, without raising implementation costs for 
scutters.

So, I'm totally sold on something like foaf:FOAFDocument as a class, where
it is a subClassOf the much more generic foaf:Document, restricted in that
it is a 'FOAF document', ie. an RDF(/XML) document that typically uses the 
FOAF RDF vocabulary.

It's the relationship between these and people that seems to require 
more thought. Eikeon asked why I wanted to assert that such a 
relationship should imply some authorship or control of the FOAF doc by
the person. I guess the main answer is: to make this something distinctly
more valuable than a 'meer' seeAlso.

Ah maybe this is the issue?: my own FOAF files all include stuff like

<foaf:Person>
  <foaf:name>Dan Brickley</foaf:name>
  <!-- blah blah etc -->
  <rdfs:seeAlso rdf:resource="thisdoc.rdf"/>
</foaf:Person>

...ie. i take care to assert that I'm described by each of my .rdf FOAF 
files, even within those files. So for me it is relatively easy to 
list all the (FOAF, if we add the new type) files that are about me.

Am I the only one seeAlso'ing my own files? 

Dan



More information about the foaf-dev mailing list