I am brainstorming strategies to generate an audio_summary for an artist.
The summary will be a function of the individual audio_summaries of each of that artist's songs. The song's audio_summary should be weighted by that song's popularity.
Out of curiosity, has anyone built something similar, and if so, what did your function look like? Would love to hear some ideas, and am open to collaborating offline if you are also building something like this.
I think this could be a useful feature to add to EchoNest in general.