W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > May 2005

GRDDL-enabling microformats

From: Danny Ayers <danny.ayers@gmail.com>
Date: Fri, 20 May 2005 11:52:03 +0200
Message-ID: <1f2ed5cd05052002525b8f8145@mail.gmail.com>
To: public-rdf-in-xhtml-tf@w3.org
Cc: tantek@technorati.com

I'm trying to get to the bottom of what is required of a microformat
publisher when they wish to make their documents available to Semantic
Web processors, and hopefully encourage Tantek to do just that...

If I understand the new GRDDL draft [1] correctly, a compatible
document would firstly include a reference to the data-view profile to
say "transformations of this doc have the same meaning as this doc" -

<head profile="http://www.w3.org/2003/g/data-view"> 

This would be the same for all microformats.

There would then follow a specific transformation link to provide a
stylesheet for the specific doc format, for example with the
DC-extractor:

<link rel="transformation"
href="http://www.w3.org/2000/06/dc-extract/dc-extract.xsl" />

This would vary per format.

Ok so far?

So perhaps the optimal action of a publisher of a microformat
specification would be to build the profile line into the spec (as a
SHOULD?) with a recommended xxx2rdfxml stylesheet for the particular
format (as a MAY?).

There is one pragmatic aspect that doesn't appear to be covered by the
microformats or GRDDL so far. Ok, a generic RDF doc might contain info
about any resources of any kind. But the microformat docs draw on a
small range of vocabularies, so I'm wondering if there may be some way
of unambiguously determining that a document contains a particular
kind of data - "this homepage contains XFN" or "this is a review".
Right now the alternatives would be either scraping the attributes to
look for some specific value (class="hreview" or whatever), or running
the whole thing through the XSLT, then querying the resultant RDF. The
former is unreliable, the latter a lot of work when most docs wouldn't
contain anything of interest.

Might it be feasible to include a big hint, something like:

<meta name="parseProfile" content="http://gmpg.org/xfn/" />

This would also help considerably when it came to data discovery - for
example if I wanted reviews of a specific book I could pass a query to
Google containing the Amazon URI of the book along with
"http://gmpg.org/hreview/"

btw, I've got a stylesheet/vocab mostly done for hReview, see [2].

Cheers,
Danny.

[1] http://www.w3.org/TeamSubmission/2005/SUBM-grddl-20050516/
[2] http://pragmatron.org/docs/misc/hreview.html 


-- 

http://dannyayers.com
Received on Friday, 20 May 2005 09:52:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:14:59 GMT