- From: Sandro Hawke <sandro@w3.org>
- Date: Wed, 21 Jul 2004 11:50:34 -0400
- To: Eric Jain <Eric.Jain@isb-sib.ch>
- Cc: rdf-interest <www-rdf-interest@w3.org>
> Tired of having nothing to test your sophisticated, high-performance > semantic web application with, except perhaps some sociological data? :-) > Get some real data from: > > http://www.isb-sib.ch/~ejain/rdf/data/ > > Feedback on the project itself is also welcome, of course :-) Excellent, now I can design some new life forms before dinner, if I can just get the inference rules right. Seriously, the data looks very cool (although it's completely outside my expertise), but you seem to have packaged it in a sort of traditional database style, intead of publishing it in a web style. Why not make your terms dereferenceable? If I understand right, you have unique/stable numbers for things (the "ac number"), so why not turn those into HTTP URIs which offer the appropriate chunks of data? The basic idea is the same as you're already using [1]; why not extend it to cover the RDF data? I think that would allow systems to use relevant parts of the data without downloading the whole thing. -- sandro [1] http://www.pir.uniprot.org/start/faq.shtml#uniplink
Received on Wednesday, 21 July 2004 11:49:13 UTC