IIIA - CSIC        Taking        people       back into   social Web music              Claudio Baccigalupo – April 2009
Timeline and motivationPoolcasting Web Radio (30’)MySpace Robot (15’)Q&A and Demo (15’)             “   e beauty of the In...
A social music experience
“Share” a radio channel?AuthoritativeWeb Radios
“Share” a radio channel?Authoritative                 PersonalisedWeb Radios                   Recommenders
“Share” a radio channel?Authoritative                     PersonalisedWeb Radios                       Recommenders       ...
POOLCASTING WEB RADIO
What is Poolcasting?
A Poolcasting radio channel
Listeners can play music
Listeners can create public channels
Participants contribute with own music
Listeners can meet other listeners
Listeners influence the music played
How to satisfy a group of listeners?         Varietythe same song or songs by the  same artist should not berepeated close...
How to satisfy a group of listeners?         Variety                      Smoothnessthe same song or songs by the   each s...
How to satisfy a group of listeners?         Variety                      Smoothnessthe same song or songs by the   each s...
How to satisfy a group of listeners?         Variety                      Smoothnessthe same song or songs by the   each s...
How to fulfil the required properties?         Variety                       Smoothness                                 ea...
How to fulfil the required properties?         Variety                      Smoothnessexclude from the channel any     whi...
Knowledge about musical associationsPoolcasting extracts this knowledge from a set of 993,825playlists compiled by the use...
Knowledge about musical associationsDRG layout of co-occurrences of songs in a set of 993,825 MusicStrands playlists
Knowledge about musical associationsDRG layout of co-occurrences of songs in a set of 106,144 Last.fm playlists
Knowledge about musical associations    Top associated songs for Smoke on the Water (Deep Purple): Space Truckin’ (VV.AA.)...
How to fulfil the required properties?         Variety                      Smoothnessexclude from the channel any     whi...
How to fulfil the required properties?         Variety                      Smoothnessexclude from the channel any     whi...
Knowledge about musical preferencesExplicit preferences for songs played or scheduled on a radiochannel can be stated usin...
Knowledge about musical preferencesImplicit preferences of participants for songs in their sharedlibraries can be inferred...
How to fulfil the required properties?         Variety                      Smoothnessexclude from the channel any     whi...
Variety                      Smoothnessexclude from the channel any      which songs and artists arerecently played song o...
The music selection algorithmShared Libraries                            Rock Channel                                     ...
The music selection algorithmShared Libraries    Music Pool   Channel Pool (Rock)   Rock Channel                          ...
The music selection algorithmShared Libraries     Music Pool       Channel Pool (Rock)   Rock Channel                     ...
The music selection algorithmShared Libraries     Music Pool       Channel Pool (Rock)     Rock Channel                   ...
Preference aggregation    ree listeners have diverging individual preferences over   which song to play after Woman in Cha...
Preference aggregation    ree listeners have diverging individual preferences over   which song to play after Woman in Cha...
Preference aggregation  To avoid misery, candidate songs that any listener “dislikes”     automatically get the lowest gro...
Preference aggregation  To avoid misery, candidate songs that any listener “dislikes”     automatically get the lowest gro...
Preference aggregation   To ensure fairness, the group preference for the remainingcandidates equals to the average of the...
Preference aggregation         e highest ranked song is selected to be played next,         leaving some listeners more sa...
Preference aggregation         e highest ranked song is selected to be played next,         leaving some listeners more sa...
Preference aggregation    To achieve fairness in the long run, the preferences of less satis ed listeners have more in uen...
Scalability of satisfaction
Scalability of satisfaction
The music selection algorithmShared Libraries     Music Pool       Channel Pool (Rock)     Rock Channel                   ...
The music selection algorithmShared Libraries     Music Pool       Channel Pool (Rock)     Rock Channel                   ...
The Poolcasting architectureplaylists                                                                  Database           ...
In favour of the “long tail” of musicEach channel plays a group-customised ordered sequence ofsongs, adapting in real time...
The future of Poolcasting
MYSPACE ROBOT
Friends as associated musicians
Identifying the most associated artists
Identifying the most associated artists
2,123 in common         Expanding the tree of friends                       Bob                      Dylan           Miles...
2,123 in common         Expanding the tree of friends                       Bob       Bob Dylan has 189,037 friends,      ...
2,123 in common         Expanding the tree of friends                       Bob     Coleman Hawkins has 261 friends,      ...
2,123 in common                 Counting shared connections                                Bob                            ...
Comparing with Last.fm similar artists
Comparing with Last.fm similar artists
Evaluating music discovery
Q&A AND DEMOhttp://www.iiia.csic.es/~claudio  http://github.com/claudiob
Upcoming SlideShare
Loading in...5
×

social web music

1,462
-1

Published on

Slides for the Poolcasting Web Radio and MySpace Robot presentation at Last.fm offices in London, April 2009

Published in: Entertainment & Humor
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
1,462
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
6
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

social web music

  1. 1. IIIA - CSIC Taking people back into social Web music Claudio Baccigalupo – April 2009
  2. 2. Timeline and motivationPoolcasting Web Radio (30’)MySpace Robot (15’)Q&A and Demo (15’) “ e beauty of the Internet is that it connects people. e value is in the other people. If we start to believe that the Internet itself is an entity that has something to say, we’re devaluing those people and making ourselves into idiots.” – Jaron Lanier (Computer scientist, composer, visual artist)
  3. 3. A social music experience
  4. 4. “Share” a radio channel?AuthoritativeWeb Radios
  5. 5. “Share” a radio channel?Authoritative PersonalisedWeb Radios Recommenders
  6. 6. “Share” a radio channel?Authoritative PersonalisedWeb Radios Recommenders Group-customised Web radio channels
  7. 7. POOLCASTING WEB RADIO
  8. 8. What is Poolcasting?
  9. 9. A Poolcasting radio channel
  10. 10. Listeners can play music
  11. 11. Listeners can create public channels
  12. 12. Participants contribute with own music
  13. 13. Listeners can meet other listeners
  14. 14. Listeners influence the music played
  15. 15. How to satisfy a group of listeners? Varietythe same song or songs by the same artist should not berepeated closely on a channel
  16. 16. How to satisfy a group of listeners? Variety Smoothnessthe same song or songs by the each song should be musically same artist should not be associated with the previousrepeated closely on a channel song played in the channel
  17. 17. How to satisfy a group of listeners? Variety Smoothnessthe same song or songs by the each song should be musically same artist should not be associated with the previousrepeated closely on a channel song played in the channel Customisationeach song should match the musical preferences of the current listeners
  18. 18. How to satisfy a group of listeners? Variety Smoothnessthe same song or songs by the each song should be musically same artist should not be associated with the previousrepeated closely on a channel song played in the channel Customisation Fairnesseach song should match the all the listeners of a channel musical preferences of the should equally have an current listeners enjoyable music experience
  19. 19. How to fulfil the required properties? Variety Smoothness each song should be musicallyexclude from the channel any associated with the previousrecently played song or artist song played in the channel Customisation Fairnesseach song should match the all the listeners of a channel musical preferences of the should equally have an current listeners enjoyable music experience
  20. 20. How to fulfil the required properties? Variety Smoothnessexclude from the channel any which songs and artists arerecently played song or artist “musically associated”? How to automatically acquire knowledge about musical associations? Customisation Fairnesseach song should match the all the listeners of a channel musical preferences of the should equally have an current listeners enjoyable music experience
  21. 21. Knowledge about musical associationsPoolcasting extracts this knowledge from a set of 993,825playlists compiled by the users of MusicStrandsPlaylists are sequences of songs ordered according tomusical, social and cultural criteria that are not discoverablewith acoustic-based analysis e more the playlists where two songs or artists co-occur,the smaller the distance at which they occur, and the smallerthe number of playlists where only one of the two occurs, thehigher their musical association
  22. 22. Knowledge about musical associationsDRG layout of co-occurrences of songs in a set of 993,825 MusicStrands playlists
  23. 23. Knowledge about musical associationsDRG layout of co-occurrences of songs in a set of 106,144 Last.fm playlists
  24. 24. Knowledge about musical associations Top associated songs for Smoke on the Water (Deep Purple): Space Truckin’ (VV.AA.) Cold Metal (Iggy Pop) Iron Man (Black Sabbath) China Groove ( eDoobie Brothers) Crossroads (E. Clapton) Sunshine of your love (Cream) Wild ing (J. Hendrix) Top associated artists for ABBA: Agnetha Fältskog A-Teens Chic Gloria Gaynor e 5th Dimension Andy Gibb Similar artists for ABBA (Last.fm): Agnetha Fältskog Frida Boney M. Bee Gees Olivia Newton-John Baccara Similar artists for ABBA (All Music Guide): Ace of Base Gemini Maywood Bananarama Lisa Stans eld Gary Wright Roxette
  25. 25. How to fulfil the required properties? Variety Smoothnessexclude from the channel any which songs and artists arerecently played song or artist “musically associated”? Customisation Fairnesseach song should match the all the listeners of a channel musical preferences of the should equally have an current listeners enjoyable music experience
  26. 26. How to fulfil the required properties? Variety Smoothnessexclude from the channel any which songs and artists arerecently played song or artist “musically associated”? How to automatically acquire knowledge about the musical preferences of the listeners? Customisation Fairness all the listeners of a channel which songs and artists the should equally have anaudience would like to hear? enjoyable music experience
  27. 27. Knowledge about musical preferencesExplicit preferences for songs played or scheduled on a radiochannel can be stated using the Web interface
  28. 28. Knowledge about musical preferencesImplicit preferences of participants for songs in their sharedlibraries can be inferred combining rating and play count
  29. 29. How to fulfil the required properties? Variety Smoothnessexclude from the channel any which songs and artists arerecently played song or artist “musically associated”? Customisation Fairness all the listeners of a channel which songs and artists the should equally have anaudience would like to hear? enjoyable music experience
  30. 30. Variety Smoothnessexclude from the channel any which songs and artists arerecently played song or artist “musically associated”?How to aggregate multiple preferences over time and satisfy all the listeners of a channel? Customisation Fairness which songs and artists the how to create a musicalaudience would like to hear? sequence that everyone likes?
  31. 31. The music selection algorithmShared Libraries Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains (Tears For Fears) ? Participants
  32. 32. The music selection algorithmShared Libraries Music Pool Channel Pool (Rock) Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains (Tears For Fears) ? Participants
  33. 33. The music selection algorithmShared Libraries Music Pool Channel Pool (Rock) Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains Retrieve candidate songs (Tears For Fears) musically associated with the last song played ? Participants
  34. 34. The music selection algorithmShared Libraries Music Pool Channel Pool (Rock) Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains Retrieve candidate songs (Tears For Fears) musically associated with the last song played Best ranked candidate Musical preferences among current listeners Participants
  35. 35. Preference aggregation ree listeners have diverging individual preferences over which song to play after Woman in Chains (Tears For Fears) possible aggregatedcandidates preferences -0.6 0.2 1 ? 0 0.2 0.4 ? 0.8 0.4 0 ? 0.2 0.6 -1 ?
  36. 36. Preference aggregation ree listeners have diverging individual preferences over which song to play after Woman in Chains (Tears For Fears) possible aggregatedcandidates preferences -0.6 0.2 1 ? 0 0.2 0.4 ? 0.8 0.4 0 ? 0.2 0.6 -1 ?
  37. 37. Preference aggregation To avoid misery, candidate songs that any listener “dislikes” automatically get the lowest group preference degree possible aggregatedcandidates preferences -0.6 0.2 1 ? 0 0.2 0.4 ? 0.8 0.4 0 ? 0.2 0.6 -1 -1
  38. 38. Preference aggregation To avoid misery, candidate songs that any listener “dislikes” automatically get the lowest group preference degree possible aggregatedcandidates preferences -0.6 0.2 1 ? 0 0.2 0.4 ? 0.8 0.4 0 ? 0.2 0.6 -1 -1
  39. 39. Preference aggregation To ensure fairness, the group preference for the remainingcandidates equals to the average of the individual preferences possible aggregatedcandidates preferences -0.6 0.2 1 0.2 0 0.2 0.4 0.2 0.8 0.4 0 0.4 0.2 0.6 -1 -1
  40. 40. Preference aggregation e highest ranked song is selected to be played next, leaving some listeners more satis ed than others possible aggregatedcandidates preferences 0.2 0.2 0.8 0.4 0 0.4 -1
  41. 41. Preference aggregation e highest ranked song is selected to be played next, leaving some listeners more satis ed than others possible aggregatedcandidates preferences 0.2 very quite not satis ed satis ed satis ed 0.2 0.8 0.4 0 0.4 ✓ -1
  42. 42. Preference aggregation To achieve fairness in the long run, the preferences of less satis ed listeners have more in uence to select the next songsuccessive possible aggregatedcandidates preferences 1 0.3 0.6 ? 0 0.9 0.2 ? 0.6 0.4 -0.1 ? -0.8 0.3 -1 ?
  43. 43. Scalability of satisfaction
  44. 44. Scalability of satisfaction
  45. 45. The music selection algorithmShared Libraries Music Pool Channel Pool (Rock) Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains Retrieve candidate songs (Tears For Fears) musically associated with the last song played ? Musical preferences Best ranked candidate Individual satisfactions among current listeners Participants
  46. 46. The music selection algorithmShared Libraries Music Pool Channel Pool (Rock) Rock Channel … Everybody Knows (Leonard Cohen) You’re in the air (R.E.M.) Woman in Chains Retrieve candidate songs (Tears For Fears) musically associated with the last song played Missing (Calexico) Musical preferences Best ranked candidate Individual satisfactions among current listeners Participants
  47. 47. The Poolcasting architectureplaylists Database MUSIC POOL MUSICAL ASSOCIATIONS CURRENT LISTENERSmetadata PREFERENCES CHANNELS list of list of available listeners shared songs songs knowledge to Stream Generator ratings and schedule play counts audio signal Library Parser Song Scheduler Streaming Server upload song OGG stream (256 Kbps) share library rate songs create channel MP3 stream Web Interface (64 Kbps) I N T E R N E T Media Personal Library Participant Player Participant
  48. 48. In favour of the “long tail” of musicEach channel plays a group-customised ordered sequence ofsongs, adapting in real time to the taste of a changingaudience without any e ort by the listenersChannels play di erent songs at di erent times dependingon which libraries are shared and which persons are listeningWhole libraries are exploited, not just the “top of the iceberg” ,while musical associations tend to favour uncommon songs,enabling people to discover or re-discover music e selection process is able to satisfy an heterogeneousgroup up, but only under a threshold number of listeners
  49. 49. The future of Poolcasting
  50. 50. MYSPACE ROBOT
  51. 51. Friends as associated musicians
  52. 52. Identifying the most associated artists
  53. 53. Identifying the most associated artists
  54. 54. 2,123 in common Expanding the tree of friends Bob Dylan Miles Davis has 27,973 friendsJohnny Miles Coleman Cash Davis Hawkins Hank Mobley Stan Getz
  55. 55. 2,123 in common Expanding the tree of friends Bob Bob Dylan has 189,037 friends, Dylan 824 shared with Miles DavisJakobDylan Sheryl CrowJohnny Miles Coleman Cash Davis Hawkins Matt Costa Hank Mobley Stan Getz
  56. 56. 2,123 in common Expanding the tree of friends Bob Coleman Hawkins has 261 friends, Dylan 115 shared with Miles Davis Quincy JonesJohnny Miles Coleman Cash Davis Hawkins Charlie Hank Parker Mobley Stan Getz
  57. 57. 2,123 in common Counting shared connections Bob 824 common friends (0%) Dylan286 common friends (0%) 115 common friends (44%) Johnny Miles Coleman Cash Davis Hawkins Hank 2,120 common friends (14%) 17 common friends (30%) Mobley Stan44 common friends (31%) Getz
  58. 58. Comparing with Last.fm similar artists
  59. 59. Comparing with Last.fm similar artists
  60. 60. Evaluating music discovery
  61. 61. Q&A AND DEMOhttp://www.iiia.csic.es/~claudio http://github.com/claudiob
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×