Well, now the Music Ontology begins to be in an usable state (as shown by the Jamendo, the Magnatune, but also Frederick Giasson's Musicbrainz dump, the EASAIER dump of the RSAMD HOTBED database), we have to ask ourselves the question: what next?

I'll try to post some ideas about that, and about potential applications of Music Ontology data (and in particular the Musicbrainz dump, as it will surely be a sort of hub for music-related data on the Semantic Web: it will give URIs for a number of tracks, artists, etc.).

The first thing I'd like to see would be embedding some RDF into an ID3v2 tag. Basically, I'd just like to put one single RDF statement:

<Musicbrainz track URI> mo:availableAs <>.

Well, that's not much, isn't it? Now, a Semantic-Web-enabled music player could follow this link, and get access to all information available on the Semantic Web, live... A good thing could be to embed the Tabulator into Songbird, then allowing you to browse the web of data from your particular item (your audio file) in your collection. Then, here are the free lunches you may get (just trying to think of funny applications - this could evolve) :

  • Place your collection on a map, according to the publication location of your track, or the performance location, or the composition location

My jamendo artists on a map

  • Generate playlists from a particular location
  • Place your collection on a timeline

My jamendo artists on a timeline

  • Generate playlists according to the composition date
  • Explore relationships between artists, generate playlists according to such relationships

Relationships between Metallica and Megadeth

Well, I hope these few examples demonstrate what can be done by interlinking geonames, dbpedia , musicbrainz and jamendo!!