Way Cool Music Search

I found Grandaddy by using Amazon's collaborative filtering technology – when I bought a Flaming Lips album on the advice of a friend, the Amazon filter said "folks who bought 'Yoshimi…' also bought…" I bit, and am glad I did. Now, take this idea to search, at least in a…

I found Grandaddy by using Amazon’s collaborative filtering technology – when I bought a Flaming Lips album on the advice of a friend, the Amazon filter said “folks who bought ‘Yoshimi…’ also bought…” I bit, and am glad I did. Now, take this idea to search, at least in a way. Thanks to Scoble for this gem: MusicPlasma. More proof of what can be done on top of a search platform. This takes a Grokker-like interface to show how your musical tastes relate. Type in one band you love and it will show you others you’d like. You can drill down in sort of a Venn diagram-driven search – say “Radiohead” then “The Shins”. As far as I can tell, this is built on the Amazon API (perhaps someone more astute than I can figure this out?). This is why, as I will say again, Yahoo and Google should really rev up their API programs.

Author: John Battelle

A founder of NewCo (current CEO), sovrn (Chair), Federated Media, Web 2 Summit, The Industry Standard, Wired. Author, investor, board member (Acxiom, Sovrn, NewCo), bike rider, yoga practitioner.

6 thoughts on “Way Cool Music Search”

  1. geez, you coulda just asked me. ;P

    there is a layer of editorial/human filtering that’s really necessary (and missing) in most of this, which is why Amazon’s muliple filtering strategies are so great.

    In examining a “The Shins” search, for example, it’d be nice to know that Doug Martsch is the frontman for Built to Spill (admittedly a close pair on the graph), or that Ugly Casanova is a Modest Mouse side project (not so close on graph).

  2. Maybe this kind of tools if good for dummy discovery, or exploring a fresh domain. But the Amazon DB is far from being complete in music, or if you really want to harvest a musical domain.
    I wish all music retailers would use a kind of Amazon API. We would be able to share DB-sources, and, most important, Amazon wouldn’t say that I could love a product I already own, bought at the Fnac.com for example.
    Personally, I discovered the Flaming Lips via Russell Beattie Notebook, blogs are everywhere 😉

  3. What about a combination of Amazon, MusicBrainz, FreeDB, and user’s would upload the metadata of music they own, this would make a nice semantic soup, I meant semantic web.

  4. The ultimate music database is over at allmusic.com. Related artists, influences/followers, where born? where died? when formed? ratings, bios, reviews, cover art, and streams. It is so complete… I would not be surprised if musicplasma was based on the allmusic.com database.

Leave a Reply