Yesterday, at the SanFran MusicTech Summit, I gave a sneak preview that showed how Spotify is tapping into the Echo Nest platform to help their listeners explore for and discover new music. I must say that I am pretty excited about this. Anyone who has read this blog and its previous incarnation as ‘Duke Listens!’ knows that I am a long time enthusiast of Spotify (both the application and the team). I first blogged about Spotify way back in January of 2007 while they were still in stealth mode. I blogged about the Spotify haircuts, and their serious demeanor:
I blogged about the Spotify application when it was released to private beta: Woah – Spotify is pretty cool, and continued to blog about them every time they added another cool feature.
I’ve been a daily user of Spotify for 18 months now. It is one of my favorite ways to listen to music on my computer. It gives me access to just about any song that I’d like to hear (with a few notable exceptions – still no Beatles for instance).
It is clear to anyone who uses Spotify for a few hours that having access to millions and millions of songs can be a bit daunting. With so many artists and songs to chose from, it can be hard to decide what to listen to – Barry Schwartz calls this the Paradox of Choice – he says too many options can be confusing and can create anxiety in a consumer. The folks at Spotify understand this. From the start they’ve been building tools to help make it easier for listeners to find music. For instance, they allow you to easily share playlists with your friends. I can create a music inbox playlist that any Spotify user can add music to. If I give the URL to my friends (or to my blog readers) they can add music that they think I should listen to.
Now with the Spotify / Echo Nest connection, Spotify is going one step further in helping their listeners deal with the paradox of choice. They are providing tools to make it easier for people to explore for and discover new music. The first way that Spotify is tapping in to the Echo Nest platform is very simple, and intuitive. Right click on a playlist, and select ‘Extend Playlist’. When you do that, the playlist will automatically be extended with songs that fit in well with songs that are already in the playlist. Here’s an example:
So how is this different from any other music recommender? Well, there are a number of things going on here. First of all, most music recommenders rely on collaborative filtering (a.k.a. the wisdom of the crowds), to recommend music. This type of music recommendation works great for popular and familiar artists recommendations … if you like the Beatles, you may indeed like the Rolling Stones. But Collaborative Filtering (CF) based recommendations don’t work well when trying to recommend music at the track level. The data is often just to sparse to make recommendations. The wisdom of the crowds model fails when there is no crowd. When one is dealing with a Spotify-sized music collection of many millions of songs, there just isn’t enough user data to give effective recommendations for all of the tracks. The result is that popular tracks get recommended quite often, while less well known music is ignored. To deal with this problem many CF-based recommenders will rely on artist similarity and then select tracks at random from the set of similar artists. This approach doesn’t always work so well, especially if you are trying to make playlists with the recommender. For example, you may want a playlist of acoustic power ballads by hair metal bands of the 80s. You could seed the playlist with a song like Mötley Crüe’s Home Sweet Home, and expect to get similar power ballads, but instead you’d find your playlist populated with standard glam metal fair, with only a random chance that you’d have other acoustic power ballads. There are a boatload of other issues with wisdom of the crowds recommendations – I’ve written about them previously, suffice it to say that it is a challenge to get a CF-based recommender to give you good track-level recommendations.
The Echo Nest platform takes a different approach to track-level recommendation. Here’s what we do:
- Read and understand what people are saying about music – we crawl every corner of the web and read every news article, blog post, music review and web page for every artist, album and track. We apply statistical and natural language processing to extract meaning from all of these words. This gives us a broad and deep understanding of the global online conversation about music
- Listen to all music – we apply signal processing and machine learning algorithms to audio to extract a number perceptual features about music. For every song, we learn a wide variety of attributes about the song including the timbre, song structure, tempo, time signature, key, loudness and so on. We know, for instance, where every drum beat falls in Kashmir, and where the guitar solo starts in Starship Trooper.
- We combine this understanding of what people are saying about music and our understanding of what the music sounds like to build a model that can relate the two – to give us a better way of modeling a listeners reaction to music. There’s some pretty hardcore science and math here. If you are interested in the gory details, I suggest that you read Brian’s Thesis: Learning the meaning of music.
What this all means is that with the Echo Nest platform, if you want to make a playlist of acoustic hair metal power ballads, we’ll be able to do that – we know who the hair metal bands are, and we know what a power ballad sounds like. And since we don’t rely on the wisdom of the crowds for recommendation we can avoid some of the nasty problems that collaborative filtering can lead to. I think that when people get a chance to play with the ‘Extend Playlist’ feature they’ll be happy with the listening experience.
It was great fun giving the Spotify demo at the SanFran MusicTech Summit. Even though Spotify is not available here in the U.S., the buzz that is occuring in Europe around Spotify is leaking across the ocean. When I announced that Spotify would be using the Echo Nest, there’s was an audible gasp from the audience. Some people were seeing Spotify for the first time, but everyone knew about it. It was great to be able to show Spotify using the Echo Nest. This demo was just a sneak preview. I expect there will be lots more interestings to come. Stay tuned.
#1 by Elias on May 20, 2009 - 5:46 am
Congratulations! I’m looking forward to try it out! :-)
#2 by Philippe Astor on May 20, 2009 - 11:10 am
Well, I’ve just one question, as an addicted Spotify subscriber in France. When will we have access to the Extend Playlist feature here ?
#3 by debcha on May 20, 2009 - 3:21 pm
So how do the rest of us in the States get the magic pixie dust to play with Spotify? Is there a timeline for a US release?
#4 by jeremy on May 21, 2009 - 8:47 pm
I hear what you’re saying with the Paradox of Choice, but my feeling is that it’s not really applicable to music recommendation in the same way.
In Barry Schwartz’s work, the paradox (inasmuch as I understand it) applies to the case where you’re really deciding between a number of items, and only able to really select one thing in the end. Which movie are you going to rent, which jar of jelly are you going to buy, which car are you going to purchase?
In those cases, you are really only interested in taking a single action: One (maybe two) movie(s), one (maybe two) jellies, and one car. The cost of a wrong decision is 2 hours (movie), 2 weeks (jelly) or 5 years (car).
Because the cost of a wrong decision is so high, too many choices does become a problem, hence the paradox that too many options becomes overwhelming.
But with music on the other hand, your goal is exploration and discovery. The cost of a wrong decision is 3-4 minutes at most, or even as short as 20 seconds if you quickly skip to the next Spotify-streamed track.
Because the cost of a wrong decision is so low, the user is more interested in listening to (and exploring) as many songs as possible. In a single 8-hour work day, you can easily get through 160 3-minute songs.
For this reason, it is my strong opinion that the Paradox of Choice does not apply. Because with music, you’re not making a choice *between* costly items, e.g. one jar of jelly that you’ll have to live with for two weeks, or one car that you’ll have to live with for 5 years. Instead, you are consuming hundreds of things per day, thousands of things per week. The paradox of choice no longer burdens you, because in effect you can choose *everything*. Or at least orders of magnitude more things than you can, when consuming cars or jelly.
So I welcome more complicated interfaces in music discovery. Music Explaura, steerable recommendations.. those things are all filled with loads of choices.. certainly way more choices than you have than in something like Google. But I don’t think Explaura etc. are burdened by the paradox of choice, even though they offer thousands of more choices, because of the nature of the consumption. The listener isn’t consuming one item; the listener is consuming many.
#5 by plamere on May 21, 2009 - 10:00 pm
Jeremy – For about a year I subscribed to eMusic – they gave me 15 mp3 downloads per month. I found the experience quite frustrating. I found it to be very difficult to find 15 new tracks every month from the millions that they offered. It was just too much work for me to decide every month. That’s why I gave up. This recent ted talk:
talks more about the difficulties in decision making and how irratiional it be, and how we can be helped by decoys.
#6 by jeremy on May 22, 2009 - 2:17 pm
Just to be clear: eMusic gave you no way of exploring the collection, other than song-by-song? So you had millions of choices, because there were millions of songs.
I agreee. That’s too much choice, and too similar to Google. It is too unwieldy to deal with all 1.3 million results in a ranked list:
But the solution is not to overreact in the other directory, i.e. the solution is not to give the user no choice (or only a single choice, which is in effect no choice) by only letting them expand the current playlist, with no input as to how that playlist is actually expanded.
What one should really have is a middle ground, where you have more than 1 choice, and less than 1.3 million choices.
And so the notion I am reacting against is the notion that having 30 choices is a “paradox of choice” in music, because it’s really not. Everyone says it is. I disagree. Having 1.3 million is overwhelming, I agree. But 30 is not, because you can easily go through 30 choices within an hour or two.
#7 by strontype on May 22, 2009 - 2:23 pm
Sorry, typo: “The solution is not to overreact in the other *direction*”, not the other directory :-)
#8 by jeremy on May 22, 2009 - 3:00 pm
Ok, I just finished watching that whole TED talk. And I agree that decision making can be difficult, and that (depending on how the question is phrased) we can easily be mislead/illusioned into going a certain direction.
I still don’t think it fully applies here, though, because the speaker is talking about making decisions between various options. You have a couple of options and you can only select one of them. Once you’ve chosen Rome, you can’t choose Paris. Once you’ve chosen Tom, you can’t choose Jerry.
With music recommendation, you’re not irrevocably choosing as much as you are exploring (”Music Explaura” :-). By it’s very nature, exploration is a very different process than choosing. In exploration, you want options. You don’t want to take a single path. You might not know very well how best to make use of those options, I agree. And 50 options is probably better than 1.3 million options. But 50 options is also much better than 1 or 2 options.
#9 by jeremy on May 22, 2009 - 3:49 pm
Sorry, this is a topic I’m passionate about, so one more link, if you’re interested:
#10 by Andreas on October 23, 2009 - 4:20 pm
Still no Echo Nest playlist feature available in Spotify client! How’s the progress here, when will it be available finally?
#11 by Eric on November 5, 2009 - 4:06 pm
+1, when are we gonna see this in Spotify?!
#12 by cuu508 on January 8, 2010 - 1:43 pm
+1, please please