FOAF-ing the music, RDF templates and suggesting content

Oscar Celma introduces a service that includes music interests in your FOAF profile. Instead of writing and updating them yourself regularly, the script will search (using SPARQL) in your profile for websites that share playlists (last.fm, pandora …) and create an updated FOAF file with your interests. I really like this approach of decentralized content to create RDF files … let computers do the job for you !

Yet I’m wondering if the script can have an option to return only the portion of RDF that should be included. It makes me think we can have "templated-RDF" that will query services to be constructed / updated. Using Oscar’s service, I can include this in my profile:

 <tpl:hasService>
  <tpl:Service rdf:resource="http://foafing-the-music.iua.upf.edu/RDFize/me">
    <tpl:hasParameter>
      <tpl:Parameter>
        <tpl:has_name>foaf</tpl:has_name>
        <tpl:has_value>http://apassant.net/foaf.rdf</tpl:has_value>
      </tpl:Parameter>
    </tpl:hasParameter>
    <tpl:hasParameter>
      <tpl:Parameter>
        <tpl:has_name>account</tpl:has_name>
        <tpl:has_value>http://last.fm</tpl:has_value
      </tpl:Parameter>
    </tpl:hasParameter>
  </tpl:Service>
</tpl:hasService> 

Then, a parser will interpret the file to add /replace this part by the resulting RDF from the service. And then, any service that return RDF could be included. Do someone know if such features already exist ?

In the meantime, I also discovered the Music Use Case, from the same team. It briefly shows how, using music metadata and people interest, it helps to suggest relevant artist and concerts, as last.fm does, but using Semantic Web technologies. This is also something I like in Semantic Web, metadata, ontologies and linked data, and it makes me think to an approach I’ll present at ICWSM, regarding blogs and information retrieval.