Live data from Hacker News

Spotify’s Discover Weekly: How machine learning finds new music

hackernoon.com

251–260 of 275 posts

Re: Spotify’s Discover Weekly: How machine learning finds new music

#251
post #145

Earlier quoted context omitted.

I gave this trick a try, but only "The Sound of ..." playlists appear in my searches.

Usually on the "Sound of.." playlist there is a link to the Pulse and Edge in the description. Otherwise the genre might be too obscure and a playlist for it not generated. Also, the "Sound of" playlists are stored under the "Sounds of Spotify" account, while the pulse and edge under "Particle Detector", so you might try searching directly in that profile.

Ok, I see the playlists now. Strong vouch for this tip!

Re: Spotify’s Discover Weekly: How machine learning finds new music

#252

Earlier quoted context omitted.

I have three main criticisms of Spotify, just in case there's a Spotify PM reading. 1. I hear repeat songs in my discover all the time, this isn't ideal. I curate my own playlists for moods and know that i'm getting when I play those playlists, surfacing songs I've manually curated in my discover doesn't add value to my experience. 2. I wish I could more easily surface the discover playlists of my friends or those wh…

>1. I hear repeat songs in my discover all the time Something is wrong. Discover Weekly is supposed to have only songs you have never listened to, even once. That's how it works for me.

The case I most often encounter is "oh, you liked x? I'm sure you'll like every remix of x for the next month of your discovery weekly!". Technically not the same song.

Re: Spotify’s Discover Weekly: How machine learning finds new music

#253

I built the foundation of this system while at Spotify. While it's true that we looked at a lot of different signal, at the point when I left (early 2015), it was all based on collaborative filtering. The reason collaborative filtering works so much better than anything else is that given enough data, it will already encompass everything else. If there are reasons why certain users prefer certain sounds, or certain l…

I guess this is a less technical question - but what would your advice be to an artist who wants fans of their more famous influences (similar artist) to discover their music? Aside from the obvious (tour, market yourself to popular playlists etc). Sounds like there's some NLP and web-scraping involved...so would it make sense to come out with blog posts that compare you to the famous artists that influence you?

only thing you can do to influence the collaborative filtering system is to make users listen to both you and some other artist.

as i mentioned in another reply, CF is really what powers DW

Re: Spotify’s Discover Weekly: How machine learning finds new music

#254

Earlier quoted context omitted.

I guess this is a less technical question - but what would your advice be to an artist who wants fans of their more famous influences (similar artist) to discover their music? Aside from the obvious (tour, market yourself to popular playlists etc). Sounds like there's some NLP and web-scraping involved...so would it make sense to come out with blog posts that compare you to the famous artists that influence you?

only thing you can do to influence the collaborative filtering system is to make users listen to both you and some other artist. as i mentioned in another reply, CF is really what powers DW

In this context, what does CF and non-CF mean?

Thanks.

Re: Spotify’s Discover Weekly: How machine learning finds new music

#255

I built the foundation of this system while at Spotify. While it's true that we looked at a lot of different signal, at the point when I left (early 2015), it was all based on collaborative filtering. The reason collaborative filtering works so much better than anything else is that given enough data, it will already encompass everything else. If there are reasons why certain users prefer certain sounds, or certain l…

Is your listening history from last.fm taken into account when signed into it via Spotify?

Re: Spotify’s Discover Weekly: How machine learning finds new music

#256
post #254

Earlier quoted context omitted.

only thing you can do to influence the collaborative filtering system is to make users listen to both you and some other artist. as i mentioned in another reply, CF is really what powers DW

In this context, what does CF and non-CF mean? Thanks.

Collaborative filtering methods vs non-collaborative-filtering recommendation methods.

Re: Spotify’s Discover Weekly: How machine learning finds new music

#258

The biggest problem with the Discover Weekly is its inability to understand _why_ you're listening to a specific subset of music. It might not be that your taste in music suddenly changed, or that you discovered a new genre that you're incredibly interested in, even though your most listened to genres or songs changed for a few weeks, or your listening patterns changed for a few weeks. A few examples: I'm Norwegian,…

We had to build some explicit filters for this at Spotify. For instance we blacklisted Christmas music from being part of the listening data – until them people would get a ton of Christmas music recommendations every January. But to your point, there's a long tail of other contextual things.

Weirdly, some genres seem to be entirely absent. I listen to a lot of jazz and classical, and with the exception of the occasional jazzy pop song, I never get recommendations for those genres.

Granted, Spotify doesn't have a huge amount of contemporary jazz (much of what I listen to are imported ECM tracks that Spotify doesn't have, though not all), but it does have a lot of classical.

I don't think it's ever recommended any Latino or folk music, either, come to think of it.

Re: Spotify’s Discover Weekly: How machine learning finds new music

#260

As an artist, Discover Weekly has been the best thing to ever happen to me. Every Monday I get a big infusion of listeners (around 5,000)— many of who stick around and check out my other music :) Prior to that, the best press I could get was the tedious process of cold-emailing bloggers (a practice which is now dying off).

How do you get onto discover weekly?
Post reply on HN