Taste Profile Subset / Song ID translation problem

Hi!

I'm using the Echo Nest Taste Profile Subset to build a small (user-based, CF) recommender system for a course of my study. I thought it would be nice to use the Echo Nest Python API to make some "real" recommendations by:

• Translating songs titles/artists to the corresponding Echo Nest IDs
• Comparing the IDs with IDs of users in the dataset, generating recommendations
• Translating the recommended IDs to song titles/artists

When translating some song titles/artists to Echo Nest IDs, I couldn't manage to find those IDs in the dataset (for 20 songs and also for a few more popular songs).

An example of a random song picked out of the dataset:

s = song.Song('SOVHRUI12AB018309A')
print(s.title)
print(s.artist\_name)

This prints the song "Moon" of the artist "Little People". All right then, let's translate it the other way around.

songs = song.search(artist = "Little People", title = "Moon")
for s in songs:
    print(s.id)

This prints SOOHRBD1311AFE26A0, SOHXGXR13D11B715A4 and SOULTME130516D9DB9. But where is SOVHRUI12AB018309A? This problem causes my recommender to return 0 similar users, because the returned IDs can't be found in the dataset...

Is there any solution to this problem?

With kind regards,

David de Kleer