With Great Honour, I Present you, Contextual Playlist in Amarok. November 6, 2009
Posted by tumaix in planetkde-tomazcanabrava, uncategorized.trackback
Given the increasing ease of obtaining all kinds of music via the Internet, the personal musical libraries are increasing more and more, so it is important to organize this collection in order to facilitate navigation and search for music throughout the music library. We propose a system for displaying music libraries in a more intuitive way for the music player Amarok, through a 2D representation (two dimensions), in order to organize the songs by the similarity between them. This representation is via contextual information allocated on special tags created by the user and associated to the musics, for example: happy, instrumental, carnival, agitated etc. or by information such as artist, album, genre, year etc. already present in the music allocated on ID3 tag.
This is the result of a job done by A group of Students, having Sandro Andrade as a Mentor and me as quick problem solver,
and since Pics or didn`t happen…
The Music Organization Dialog
Circular vizualization by similarity
Aggregation by similarity.
This is a ongoing research project, so, nothing is finished yet, but as you can see, we are pretty busy here in Brazil 😉
Rocs, KDevelop, Plasma by the INdT guys, Kopete by Lamarque, and now, Amarok, by André Simpatia & colleages.
Live Blue 
NEPOMUK integration, please 😉
(I’m raving with the alpha-quality-but-insanely-great-and-freaking-amazing NEPOMUK Smart Desktop included in Mandriva 2010.0, so excuse me for the intro :P)
Cool project, I’m really looking forward to trying it out with my music collection. =)
Awesome!
Back in my Windows days I used to use a program called MoodLogic that was similar and I *loved* it. I’m *thrilled* to see this coming to my beloved Amarok!
What version of Amarok can we expect to see this in?
Now if we could get some kind of system for everyone to share their tag ratings, then it could save people a massive amount of tagging work.
Thankyou Team FOSS Brazil! \o/
Awesome mister, simply awesome!
As you said, the sources of obtaining music are ever increasing, so are the number of tracks and artists.. you can have your personal radio solely made up from your favorite songs.
For example, every month I get huge loads of fresh music from blogs and hypem, and there are many artists with only one or very few songs.
So a linear playlist doesn’t cut it anymore when you want to keep track or rediscover them when there’s a constant flow of new stuff coming in.
At the moment it’s more like fire and forget, you archive them and they’re gone. But they are worth a second listen, and it would be great if they were at hand at the right time.
Your system of displaying and accessing the songs looks innovative and capable of addressing the problems coming with the excess supply of music.
Ideas: Automate tagging (kind of) based on where it’s coming from (ask Nepomuk etc) and other sources, say lastfm.
This is awesome mister, simply awesome!
As you said, the sources of obtaining music are ever increasing, so are the number of tracks and artists.. you can have your personal radio solely made up from your favorite songs.
For example, every month I get huge loads of fresh music from blogs and hypem, and there are many artists with only one or very few songs.
So a linear playlist doesn’t cut it anymore when you want to keep track or rediscover them when there’s a constant flow of new stuff coming in.
At the moment it’s more like fire and forget, you archive them and they’re gone. But they are worth a second listen, and it would be great if they were at hand at the right time.
Your system of displaying and accessing the songs looks innovative and capable of addressing the problems coming with the excess supply of music.
Ideas: Automate tagging (kind of) based on where it’s coming from (ask Nepomuk etc) and other sources, say lastfm.
This looks very interesting 🙂
Can you elaborate some more on how his this integrates with Amarok? E.g. is it a patch, or a script, or plugin?
It’s a plugin. being developed as a part of conclusion project from some students. so it will be avaliable as public as soon as they finish it.
Indeed Brazilian developers are very active and innovative! It’s great to see such an energy and new ideas and fast results! We had students working on KBruch and they did an amazing work as well.
Thanks to Brazil and make sure you are all very present at next Akademy!
Hugs to you all 🙂
This looks really interesting!
I am one of the Amarok developers, and we would love to hear more about how you have done this and potentially work together with you in the future if you re interested in bringing this to a wider audience.
Is there someway we can get to try this out or is that too early?
– Nikolaj
[…] With Great Honour, I Present you, Contextual Playlist in Amarok. « Live Blue a few seconds ago from Choqok […]
[…] With Great Honour, I Present you, Contextual Playlist in Amarok. « Live Blue a few seconds ago from Choqok […]
[…] With Great Honour, I Present you, Contextual Playlist in Amarok. This is the result of a job done by A group of Students, having Sandro Andrade as a Mentor and me as quick problem solver, […]
this should be integrated with last.fm and auto-fetch the last.fm tags for all the songs. nowadays, you just can’t expect the users to tag their 50+ GB libraries…
this probably will be implemented in the near future, now it’s just a proof of concept.
[…] With Great Honour, I Present you, Contextual Playlist in Amarok. « Live Blue […]
Is the plan to share this information between users? As you say music collections are typically very large and tagging every track with a dozen or so numbers by hand would be many days work for a lot of users.
in the beginning, yes. the next step is automatic tagging.
its perfect. Whatever innovation for the good – I say bring it on!!! 🙂
keep up the great work guys!!! Youre changing the world by bringing great Free/Libre applications to it! Soon, this applications will dominate over the proprietary spy-crap. Thank you for your efforts!
Have you looked at mirage? It describes itself as:
“Mirage is an implementation of the latest research in automatic playlist generation and music similarity. Mirage analyzes your music collection and computes acoustic similarity models for each song. After your music collection has been analyzed, Mirage is able to automatically generate playlists of similar music.” from http://hop.at/mirage/
Currently, it’s a plugin for Banshee, but the author has the theory available as PDF.
hi, i’am Rafael the main developer (actually the only developer ’til now) of this plugin, called YAMO. I’ve started this project with two colleages a few months ago . Before this project even started we had a hard time to find a decent idea for the conclusion work for the university . That until Sandro Andrade came up with this idea : ). So , the development of the plugin is almost finished (the 1.0 version that is), i have only to kill some kinda nasty bugs but the main functionality is there. I plan to release it soon. The development process was both a pain and a great pleasure at the same time. But it is far from finished 🙂 i have a bunch of ideas and functionality to add to YAMO so i and my teammates will keep working hard on it XD . Thank you all for the great feedback ! : )