[EP-tech] Issue in EPrints RDF URIs

[cjg at ecs.soton.ac.uk (Christopher Gutteridge)]

One other thing is a fundamental cock-up I made when I designed the URI 
schemes for linked data in EPrints. It uses a single URI to indicate 
both the catalogue record and the subject of the catalogue record. (see 
the recent post about RSS feeds "published" time). Really EPrints 
records talk about two objects; the catalogue record and the subject of 
the catalogue record, and these should have distinct URIs. My instinct 
would be to use a new URI for the catalogue records as most of the data 
is about the subject, eg. 
http://eprints.soton.ac.uk/id/eprint/50995#catalogue

In actual fact, the RDF doesn't, by default, mention any properties of 
the catalogue record, such as the deposit time, or depositing user etc., 
so this is quite easy to fix sometime in the future.

If I were implementing things today, I'd use 
https://www.w3.org/TR/vocab-dcat/ but that didn't exist in as complete a 
form back then (also it is slightly badly designed in that it talks 
about "datasets" and it's? touch-and-go if EPrints records are 
datasets... I really wish this was just a catalogue vocabulary.

-- 
Christopher Gutteridge -- http://users.ecs.soton.ac.uk/cjg

University of Southampton Open Data Service: http://data.southampton.ac.uk/
You should read our Web & Data Innovation blog: http://blogs.ecs.soton.ac.uk/webteam/