mtaste: Coordinating Musical taste
Comments: 6
Certainly lots of music-oriented posts springing up in the last couple of days. ;-) Here’s a neat idea from Steven Frank about trying to generate a sort of coordinate system for people’s musical tastes. The problem:
With the debut of Apple’s music service, I’m reading a lot of blog entries that say “Check out band so-and-so, they’re great!” But as much as I love to discover new music, I invariably don’t check out the recommended band. Apart from the fact that I’m lazy, why on earth not?
Well, because I have no way of knowing what the author’s baseline musical tastes are. You can tell me a Garth Brooks song is great, but odds are I’m not going to feel the same way. Musical tastes are not objective, and if I don’t know where you’re coming from, then it could just be time consuming to check out the recommendations. But, on the other hand, I hate to think that I might be missing out on music I would actually like.
Steven’s come up with the core of an idea that could let people work out a sort of taste baseline with regard to artist and track recommendations. Lots of problems, as he admits, but then trains had lots of problems when they started out, too, and now we ride them every day. Sounds a bit like what Firefly (was it Firefly?) was trying to do back in the mid-’90s with their recommendation system (remember the one where you had to go through and rank various things, and the more you ranked, the better — supposedly — would be the recommendations?).
•••
Posted to General Rants • 2003.04.30 (Wed) • 16:17
Comments
Posted by Tom 2003.04.30, 19:07
Sounds like what’s needed is a combination of Music Brainz and Audiscrobbler. (http://www.musicbrainz.org and http://www.audioscrobler.com). With a hint of Amazon and All Music Guide as well, of course.
Posted by M Sinclair Stevens 2003.04.30, 23:55
And Netflix does it for movies. It’s pretty impressive when you rate about 100 movies, but after that the effect seems to flatten out. The nice thing is that Netflix shows the ratings for both their total customer base and the customers whose profile is similar to yours.
My standby is still to know my reviewer. For example, in my youth I knew if Rolling Stone pooh-poohed something as “pretentious”, I’d probably love it.
Posted by Aa 2003.05.01, 10:06
http://www.gnoosic.com/
“Gnod is a self-adapting system that learns about the outer world by asking its visitors what they like and what they don’t like. In this instance of gnod all is about music. Gnod is kind of a search engine for music you don’t know about. It will ask you what music you like and then think about what you might like too. When I set gnod online its database was completely empty. Now it contains thousands of bands and quite some knowledge about who likes what. And gnod learns more every day.”
They’ve also adapted the same system to movies and books, very cool site.
Posted by WP 2003.05.01, 11:34
The concept behind Firefly is based upon something called “machine learning”. For a list of (slightly outdated) applications using these concepts check out: http://www.aic.nrl.navy.mil/~aha/research/machine-learning.html
Posted by Poochie 2003.05.02, 02:39
And then there’s article in today’s New York Times. Theres a quote in there somewhere from Pattie Maes who came up with the tech behind Firefly. Turns out Microsoft bought the software not for the recommendation aaspect but for the user id system which would become the basis of Passport.
http://www.nytimes.com/2003/05/01/technology/circuits/01reco.html
Posted by Jonny 2003.09.09, 04:17
Check out Musical Taste at: http://www.musicaltaste.com
It’s a database of individual song recommendations made by users. It gets a fair bit of use from people who search for a particular artist or song, and are then able to check out other songs that people who like that song recommended. It’s also my site. Sorry about that!
Post a comment:
Send This Story to an Enemy
• • •
