The year is about to end, and as usual Spotify has launched a Year in Music microsite for global and personal trends.

The global stats are cool, but largely unsurprising. What I really like is the breakdown of your personal listening habits by genre, artist, day of the week, season, and more. You can even see the split between desktop and mobile device streaming, and, if you’re a Premium subscriber, the location of your favorite artists on a chart.

I’m a big fan of Spotify’s use of data to improve algorithms, and this is another nice experiment. If you’re into this kind of visualizations, I also recommend the excellent Exist.io service, which supports Spotify (and many other wearable devices/apps).

As for my top songs:

Death Cab for Cutie and Tegan and Sara? Looks about right.

Spotify knows what time of day users listen to certain songs, and in many cases their location, so programmers can infer what they are probably doing—studying, exercising, driving to work. Brian Whitman, an Echo Nest co-founder, told me that programmers also hope to learn more about listeners by factoring in data such as “what the weather is like, what your relationship status is now on Facebook.” (In 2011, Facebook entered into a partnership with Spotify.) He added, “We’ve cracked the nut as far as knowing as much about the music as we possibly can automatically, and we see the next frontier as knowing as much as we possibly can about the listener.”

John Seabrook’s article on Spotify for The New Yorker is a good one. A lot of interesting details about the company’s CEO, Daniel Ek, and the way they make deals with labels.

I remember trying Spotify many years ago with a fake UK account and telling my girlfriend that it was incredible and the future of music was going to be streaming. In the years I’ve spent jumping between music streaming services, I’ve kept an eye on Spotify and their marketing efforts, which the article doesn’t mention (all my friends in Italy know what Spotify is; my mom uses it).

I’ve recently started using Spotify again as my main streaming service because of its solid iPad app and new Family accounts. I am, however, excited to see what Apple does with Beats Music. I don’t know if the entire music industry will embrace streaming eventually, but the future is definitely interesting.

From the Spotify blog:

Are you currently sharing your Spotify account with the entire family? Want to keep your 60s soul classics playlist separate from your kids’ Frozen soundtrack and save money in the process?

Well, great news! With Spotify Family you can now invite up to four family members and share one billing account whilst keeping your listening history, recommendations and playlists completely separate.

Family subscriptions were one of the main reasons my girlfriend and I used Rdio with a family account. Now, we’ll be able to give Spotify a try and see if it works better for her (understandably, she doesn’t care about trying every single streaming service like I do) – and plans will also be cheaper for two users (Rdio starts at $17.99 for two members, Spotify Family will be $14.99).

Spotify Family subscriptions will be available soon.

This summer, I’m interning at Spotify in New York City, where I’m working on content-based music recommendation using convolutional neural networks. In this post, I’ll explain my approach and show some preliminary results.

A fascinating post by Sander Dieleman explaining how he designed a deep neural network to provide music recommendations based on Spotify data. It’s a technical read, but there are Spotify player widgets to listen to songs picked up by the network, which is capable of learning specific pitches, chords, sounds, and other factors such as Chinese voices. Dieleman says that hopefully some of these audio-based recommendations will go into A/B testing soon, with a focus on recommending new music that isn’t generally picked up by traditional recommendation engines. Impressive.