My problem with discover weekly is I basically listen to boring background music like vaporwave all day at work. So while I like a large variety of music from all sorts of genres, since I spent 8 hours a day listening at work, my discover weekly is just slammed full of that stuff.
Spotify’s Discover Weekly: How machine learning finds new music
101–110 of 275 posts
Re: Spotify’s Discover Weekly: How machine learning finds new music
#102If there is at least one song that I like in Discover Weekly, then I consider it to be a good mix. However, this happens once a month at best. After reading the article and all three methods for recommendations, I cannot understand how on Earth it comes up with its suggestions. Do I feed it bad data? Does it not take into account the relative "weight" of a particular track, i.e. my preference to listen to it multiple…
Yeah, I was surprised when I read the start of the article. On Mondays, I will start the day by playing my discover weekly playlist and I'll usually have it running throughout the day, saving 1-4 songs from it to my library. Sometimes none. I just checked and this week I didn't even save a single song for example. I guess it doesn't help that I do listen to really different types of music depending on what I'm doing.…
This is why they have Daily Mixes, which are clustered by genre.
Discover Weekly seems to have figured out that one of my most listened-to Daily Mixes is post-rock, so it only gives me that. I'm fine with that. I wish they would push the Daily Mixes and Discover Weekly more prominently in the app interface rather than giving me TOP 10 BRAZIL playlists when I open the app.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#103My problem with discover weekly is I basically listen to boring background music like vaporwave all day at work. So while I like a large variety of music from all sorts of genres, since I spent 8 hours a day listening at work, my discover weekly is just slammed full of that stuff.
You could turn on Private Listening[0] at work, that way the boring background music won't affect your suggestions.
[0] https://support.spotify.com/us/using_spotify/the_basics/how-...
Re: Spotify’s Discover Weekly: How machine learning finds new music
#104In case you didn't know, Spotify categorizes music into genres behind the scenes. You can use this site to find out what genres your favorite artists are categorized as: http://everynoise.com/engenremap.html You can use this information to then check out Spotify's auto-generated playlists for each genre. They have at least three types for each one: "The Sound of ", containing definitive representation of the genre, "…
> You can use this site to find out what genres your favorite artists are categorized as
That's my problem right there: my favorite song right now is a musician's piano cover of one of his own songs. His music is usually electronic, which I don't like, but I love this one song. So Spotify will recommend me electronic music from other artists, which of course does not fit my song.
Repeat for every author. I liked one song from a German musician, and now half the recommendations are German music. While I can see why network relations make sense, I wish I could say "give me a similar song, not a similar artist".
Re: Spotify’s Discover Weekly: How machine learning finds new music
#105If there is at least one song that I like in Discover Weekly, then I consider it to be a good mix. However, this happens once a month at best. After reading the article and all three methods for recommendations, I cannot understand how on Earth it comes up with its suggestions. Do I feed it bad data? Does it not take into account the relative "weight" of a particular track, i.e. my preference to listen to it multiple…
Same. After I read this article I wondered if I'm doing something wrong to generate the playlist.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#106Sometimes I wish Spotify added a bit more 'noise' to their recommendations, so to speak. If I don't listen to much music except Discover Weekly for a few weeks, I (subjectively) find that what's recommended to me more or less sounds the same after a while. Either they are afraid to insert new things that stray too far from an optimal recommendation or they forget too much of my listening history.
I used to listen exclusively to Discover Weekly for a while. One week a finish rap song made it to the list, and I skipped past it every time it came on. Next week there was 2 finish rap songs. Then 5. Eventually half of my discovery list was finish rap, something I have no interest of. Canceled my subscription shortly after. Their algorithms are feeding themselves. I wish they had a dislike button so I could at leas…
I have the exact same thing with Dutch rap music. I'm not Dutch, don't speak Dutch and have never listened to Dutch music. But Spotify consistently puts 2-4 Dutch songs on my Discover Weekly.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#107If there is at least one song that I like in Discover Weekly, then I consider it to be a good mix. However, this happens once a month at best. After reading the article and all three methods for recommendations, I cannot understand how on Earth it comes up with its suggestions. Do I feed it bad data? Does it not take into account the relative "weight" of a particular track, i.e. my preference to listen to it multiple…
It works wonders for me, to be honest. At least 50% is on spot, almost every week.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#108If there is at least one song that I like in Discover Weekly, then I consider it to be a good mix. However, this happens once a month at best. After reading the article and all three methods for recommendations, I cannot understand how on Earth it comes up with its suggestions. Do I feed it bad data? Does it not take into account the relative "weight" of a particular track, i.e. my preference to listen to it multiple…
I have at least one song I like in every Discover Weekly edition, sometimes several. However, interestingly, the overall quality and interest match varies heavily from week to week. Some weeks are fantastic, others rather lame.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#109Can't they just make great recommendations based on: People who have music X on their playlist also have Y a lot. Person A listens to X but not Y. Let's make them discover Y. You could just build a topology of songs like that and then recommend songs to user A if they are topologically close to the songs he likes. EDIT: Read the article now :D they do that and it's called Collaborative Filtering.
It's also how Criticker recommends films, and I have to say their predicted scores are uncanny. After watching a film, I often think of a 1-100 score and then look it up on their site, and it's at most ±2 from it. (Not affiliate, just a big fan)
Another movie recommendation website is MovieLens, http://movielens.umn.edu/ which is run by a research group of University of Minnesota.
Re: Spotify’s Discover Weekly: How machine learning finds new music
#110Does anyone know how they evaluate how successful their new recommendation algorithms are? How much do they improve on simpler algorithms? It's not obvious/intuitive the NLP and audio model algorithms mentioned would be that successful. I would have thought collaborative filtering + showing you new tracks from artists you like + showing you new tracks in genres you like would get you most of the way there.
I really don't understand why collaborative filtering isn't the sole approach. It's basically exactly what you want if you have that volume of data.