Old, ambient Japanese music became a smash hit. obscure chillout tunes are becoming popular again due to being featured in situational playlists.
Tag: recommendations
The Satisfaction Paradox
a future where amazon and others sell the recommendations as a service but the actuals goods are free.
While it may be a long while before every adult is sharing art or innovations on a global scale, we can already see the abundance of good stuff piling up. Netflix has more great movies a click away — after I filter out the dross — than I can watch in my lifetime. What do I watch next? Spotify and other music streaming services will have more fantastic, I-am-in-heaven music available everywhere all the time than I can ever listen to. What do I listen to next? Google will have every book ever published only 125ms away, and collaborative filtering, friends recommendations and a better Amazon engine, will narrow down those stacks to the best 10K books for me. So what do I read next?
I believe that answering this question is what outfits like Amazon will be selling in the future. For the price of a subscription you will subscribe to Amazon and have access to all the books in the world at a set price. (An individual book you want to read will be as if it was free, because it won’t cost you extra.) The same will be true of movies (Netflix), or music (iTunes or Spotify or Rhapsody.) You won’t be purchasing individual works.
Extreme Netflx
So, you think you’ve honed the Netflix recommendation engine by rating 1000 movies? That’s nothing, according to the company’s internal statistics. Several 100 Netflix members have rated more than 50k filmed entertainment programs. To watch all those at a pace of 1 movie or TV show per day, it would take 136 years. But those users are just the extreme end of a broader behavioral pattern. 0.07% of Netflix users have rated 20k items. 1% have rated 5k movies. Only 60% of Netflix users rate any movies at all, and the typical person only gives out 200 starred grades.
StyleFeeder Tech Blog
woo! cutting-edge recommendation tech
Jason Rennie
recommendations / machine learning guy (now working for phil)
Netflix Prize
Recommendation value
Personalization and recommendations apparently are responsible for 35% of sales at Amazon.com.
Amazon > Husband
Amazon has a more thorough, individualized, and nuanced understanding of Meyers’ taste than the man who occasionally claims to love her, her husband.
has last.fm jumped the shark?
last.fm used to give me great recommendations, now it gives me nirvana and 9 inch nails when i ask for bands similar to air. looks like the site is being overrun by people with horrible music tastes, alas.
Recommendations done right
i have recently started to use last.fm more frequently, again. i have had an account there since 2003, but had forgotten about it. in the meantime, they have built out an awesome service that works really well. it was a nice surprise to see that they upgraded early adopters to their version of a pro user, for free. this allows me to have my own radio station without lifting a finger, and other nice benefits.
coupled with their sane data policy, this is a clear winner. really useful and relevant unlike the overhyped silliness that is “podcasting”.