Yesterday, I put online the John Peel sessions as linked data (dereferencable identifiers, content negotiation, RDF, etc.).

It uses the data the BBC has released for the Hackday, some weeks ago. I wrote a SWI-Prolog wrapper for this data, which is then made accessible through SPARQL using P2R (which I have updated to handle dynamic construction of literals, by the way) and this mapping. The URIs are then made dereferencable through UriSpace.

Some documentation is available there.

Here are a bunch of URIs that you can try:

And then, for example

$ curl -L -H "Accept: application/rdf+xml"
<?xml version='1.0' encoding='UTF-8'?>
    <!ENTITY foaf ''>
    <!ENTITY mo ''>
    <!ENTITY rdf ''>
    <!ENTITY rdfs ''>
    <!ENTITY xsd ''>

<mo:MusicArtist rdf:about="">
  <rdfs:label rdf:datatype="&xsd;string">King Crimson</rdfs:label>
  <foaf:img rdf:resource=""/>
  <foaf:name rdf:datatype="&xsd;string">King Crimson</foaf:name>

<rdf:Description rdf:about="">
  <mo:performer rdf:resource=""/>

<rdf:Description rdf:about="">
  <mo:performer rdf:resource=""/>


So far, this dataset is not linked to anything external! But I plan to link it to Musicbrainz, Geonames, and snippets soon.