[microformats-discuss] Microformat validation

[Danny Ayers]

On 8/21/05, Bud Gibson <bud at thecommunityengine.com> wrote:
> I have to confess that most of this flew over my head.  What I think
> you are saying is something like this.
> 
> "While microformats are not optimized for machine validation out of
> the box, we might be able to get around that by associating them with
> some parallel ontology constructions that are more amenable to
> automated processing.  Then, by analogy with these parallel
> constructions, we could determine whether the page author was using
> microformat components consistent with their semantics."

Yes, very nicely put. 

> Sounds like a worthy goal, but I wonder if you might not achieve a
> similar end by dropping the middleman (RDF/OWL?) and writing the XSLT
> rules (or some imperative equivalent) directly.  

Perhaps, but I've no idea where you would start ;-) 
RDF+OWL offers a common model in which quite a range of semantics can
be expressed, it can be mapped in a straightforward fashion from
microformat data and data can be mixed from different sources without
breaking the model.

It would be a bit
> more ad hoc, but as you mention we don't have the RDF for most
> microformats and the machinery for the processing is not there yet.

Ah, sorry - I meant to say there *is* quite a lot of the RDF that
would be needed available already, and the machinery is available off
the shelf . Some of the XSLT from microformat to RDF/XML is still
lacking, as are the little bits of glue code.

I want to use microformat data within RDF systems myself so may well
end up filling in the blanks myself - but at the rate I've been going
chances are someone else will have done these bits first - hopefully
;-)

Cheers,
Danny.

-- 

http://dannyayers.com