- From: T.Heath <T.Heath@open.ac.uk>
- Date: Thu, 11 Oct 2007 16:15:43 +0100
- To: "W3C SWEO IG" <public-sweo-ig@w3.org>, <danny.ayers@gmail.com>
- Cc: <t.heath@open.ac.uk>
[moving this into my work mailbox...]
Nice one, thanks Danny.
> [cc'ing sweo]
>
> Hi Tom,
>
> I suspect it might be easiest for me to manually add the list
> of resources on the Wiki:
>
>
http://esw.w3.org/topic/SweoIG/TaskForces/InfoGathering/RecommendedTutor
ials
> what do you reckon?
I think that would be a real shame...
> - but following your suggestion of putting stuff in RDF for more
direct import,
> I've had a quick play with aggregating data from del.icio.us with an
appropriate > combination of tags ("tutorial"+"rdf"
> etc) and sanitising it. (The raw RSS/RDF from del.icio.us makes the
tagger the
> dc:creator of the resource, I think there were other questionable
statements...).
> Under http://n2.talis.com/svn/playground/danja/python/sweo-tutorials/
> you will find:
> delicious-source.py
> - quick and dirty RDF accumulator
> delicious.rdf
> - what a trial run aggregated
> delicious.sparql
> - to grab the bits of interest
> delicious-sweo.n3
> - the result
> I used rdflib to accumulate the stuff, but that doesn't yet support
SPARQL
> CONSTRUCT (hint hint Ivan!) so I used ARQ for that.
> rdflib also seems to have done something strange with the taggings in
the source > data (which uses taxo:), to the extent I couldn't easily
swap these over to the
> tag ontology - so I hard-coded a couple of taggings for all items.
Ok, excellent, I'll have a look at your output.
> btw, might just have been me, but when I checked the RDF for a recent
post at
> revyu.com I only found some garbled stuff.
Ah, ok, thanks. I'll investigate...
More soon,
Cheers,
Tom.
Received on Thursday, 11 October 2007 15:16:06 UTC