- From: Stefano Mazzocchi <stefano@apache.org>
- Date: Wed, 09 Jun 2004 05:48:05 -0400
- To: "Butler, Mark" <mark-h.butler@hp.com>, Simile <www-rdf-dspace@w3.org>
Received on Wednesday, 9 June 2004 05:48:07 UTC
Butler, Mark wrote: > Hi Stefano > > >>but in the MADS schema you get >> >> ... >> <xsd:attribute name="relatedType" use="optional"> >> ... > > >>so, if the attribute is not present, the XML is legal, but the >>stylesheet wouldn't work. > > > I hadn't intended the stylesheet to be complete, just to illustrate how you > could do this. Mark, I was not commenting on your stylesheet but on the fact that it is always going to be algorithmically uncertain to transform such MADS XML fragments into RDF statements if the type of relationship is kept optional. Of course there is lots of relationship information in the litterals that are being connected (for example from "new york" ---> "big apple" we can infer the relationship "also referred to as", but this is never a mechanical transformation). But my point is rather simple: no matter how it is encoded, the information for all three items of the statement in the MADS XMLSchema should be mandatory, or it will be painful, or heuristical, to convert it to RDF. -- Stefano.
Received on Wednesday, 9 June 2004 05:48:07 UTC