- From: Danny Ayers <Danny.Ayers@talis.com>
- Date: Mon, 15 Oct 2007 09:38:35 +0100
- To: "T.Heath" <T.Heath@open.ac.uk>, "T.Heath" <th2499@openmail.open.ac.uk>, "W3C SWEO IG" <public-sweo-ig@w3.org>, <danny.ayers@gmail.com>
> A bit of hacking on your source data has produced this: > <http://kmi.open.ac.uk/people/tom/tmp/sweo/tutorials-output.rdf> > (tragic how laborious it was; oh for wider support of SPARQL property > functions) > > Structure-wise this is pretty much ready to go into Revyu. Great! Is > there any chance of another version of the source n3 file > with a bit of additional cleaning up of the data? Yep, sure. Leave it with me. But as Leo just reminded me - don't let this interfere with your PhD work!! I think I > spotted a few versions of the Linked Data tutorial at > different URIs and some other duplication (rdf:about was one > case); there are also some meaningless titles which it would > be really good to change before importing, plus some possible > character encoding issues, and possibly some dodgy CDATA. > > What dya think? rdflib seems fairly easy-going, probably why character sloppiness is slipping through. When I get a minute I'll rework the code using Jena, I think that's considerably more anal. Not sure how best to deal with duplicates etc though. Hmm, for this RecommendedTutorials bit there really aren't *that* many resources that will need putting in the system, doing it manually may still be the best option. Longer term it would be nice to be able to automate things, but that's definitely lower priority. > In general though it's looking great, and will be a cool > addition to Revyu; hopefully Revyu can make a good > contribution in return in terms of infrastructure. Well yeah, I really like the idea of hooking it up, getting a practical application outta this linked data stuff... Cheers, Danny.
Received on Monday, 15 October 2007 08:38:54 UTC