SlideShare a Scribd company logo
1 of 25
Download to read offline
I have 10 million
   songs in my
     pocket.

    Now what?

Paul Lamere
ACM Recommender Systems 2012   Photo (CC) by Jason Rogers
The challenge of music recommendation
A recommendation that no human would make



   If you like Britney Spears ...




  You might like the Report on
      Pre-War Intelligence
The challenge of music recommendation
WTF - Why the Freakomendations?




          Why do we see such bad
          music recommendations?

              Music is Special*


                                  *of course every domain is special
The challenge of music recommendation
Why do we see such bad music recommendations?




        Understanding the domain is
         critical to the success of a
                recommender
The challenge of music recommendation
Why do we see such bad music recommendations?




       10 things to consider when
      building a music recommender
#1 - Very large item space
10 things to consider when building a music recommender
#2 - Very low cost per item
10 things to consider when building a music recommender
#2 - Very low cost per item
10 things to consider when building a music recommender
#3 - Low consumption time
10 things to consider when building a music recommender




       The average song length is around 4 minutes


                   E Book Reader:      24 books a year
                   Music subscriber:   25 songs a day



Pew Internet Project.
#4 - Highly Interactive
10 things to consider when building a music recommender


     A typical music recommender ...




    ... in 1999, but not anymore
#4 - Highly Interactive
10 things to consider when building a music recommender


                                   Today’s recommender



                                     Recommendations are
                                         integrated into
                                    the listening experience
#4 - Highly Interactive
10 things to consider when building a music recommender


                                   Today’s recommender


                                    Minimal explicit feedback
#4 - Highly Interactive
10 things to consider when building a music recommender


                                   Today’s recommender
                                   Much Implicit Feedback
                                      - Playing
                                      - Skipping
                                      - Repeating
                                      - Adjusting the volume
                                      - Sharing with friends
                                      - Adding to a playlist
                                      - Repeating the song
                                      - Inspecting song info
#5 - Very high per-item reuse
10 things to consider when building a music recommender

               songs[“as time goes by”].playcount += 1
#6 - Large personal collections
10 things to consider when building a music recommender
Personal Music Discovery Challenge
There’s a long tail in my iPod

                                          Listener Study

                                    Listeners         5,000

                                  Average Songs
                                                      3,500
                                    Per User

                                 Percent of songs
                                                      65%
                                 never listened to
#7 - Consumed in sequences
10 things to consider when building a music recommender



 A good playlist is a balance of:

     • Coherence
     • Familiarity
     • Discovery
     • Variety
     • Surprise
   ... in a pleasing order
#8 - Highly contextual usage
10 things to consider when building a music recommender
#9 - Highly passionate users
10 things to consider when building a music recommender
Let’s pause for a quiz


                            Why is this formula troublesome for
                           music recommendation and discovery?




(ΔMī¹=αΣDi[n][ΣFij[n-1]+Fexti[[n̄¹]])
Let’s pause for a quiz


                            Why is this formula troublesome for
                           music recommendation and discovery?




             Because it is the name of a
             song by Aphex Twin



(ΔMī¹=αΣDi[n][ΣFij[n-1]+Fexti[[n̄¹]])
#10 - OMG Metadata
10 things to consider when building a music recommender


  The The
  Duran Duran Duran
  !!!
  †††
  ///▲▲▲
  ▼□■□■□■
  Various Artists
#10 - OMG Metadata
10 things to consider when building a music recommender
Music Taste is Irrational
The challenge




           Music               Music
           I Like             You Like
                     Music
                     I Used
                    To Like



                                Get this t-shirt at dieselsweeties.com
I have 10 million
   songs in my
     pocket.

     Now what?

Paul Lamere
 Paul@echonest.com
 MusicMachinery.com
 @plamere

                      Photo (CC) by Jason Rogers

More Related Content

What's hot

Questionnaire analysis
Questionnaire analysisQuestionnaire analysis
Questionnaire analysisAkshata Mishra
 
Music questionnaire2015
Music questionnaire2015Music questionnaire2015
Music questionnaire2015Akshata Mishra
 
1st Draft of Power Players Magazine Editor Troy Taylor
1st Draft of Power Players Magazine Editor Troy Taylor1st Draft of Power Players Magazine Editor Troy Taylor
1st Draft of Power Players Magazine Editor Troy TaylorTroy Taylor
 
Music video questionnaire
Music video questionnaireMusic video questionnaire
Music video questionnaireCharlie12456
 
Analysis of survey results
Analysis of survey resultsAnalysis of survey results
Analysis of survey resultsthe-hunger-games
 
Questionnaire analysis for a2
Questionnaire analysis for a2Questionnaire analysis for a2
Questionnaire analysis for a229542rayner
 
Researcb audience
Researcb audienceResearcb audience
Researcb audiencetiapeek
 
'Lyric' magazine
'Lyric' magazine'Lyric' magazine
'Lyric' magazinedavid-hill
 
Contents ideas
Contents ideasContents ideas
Contents ideasramsz001
 
Interviewee powerpoint
Interviewee powerpointInterviewee powerpoint
Interviewee powerpointJulia Barfoot
 
Target audience analysis
Target audience analysisTarget audience analysis
Target audience analysisdemioshea
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire resultslucy_richmond
 
Target Audience Survey
Target Audience SurveyTarget Audience Survey
Target Audience SurveyLydiaNoble
 
Betsy Bolte Artists & Labels 2016 LinkedIn
Betsy Bolte Artists & Labels 2016 LinkedInBetsy Bolte Artists & Labels 2016 LinkedIn
Betsy Bolte Artists & Labels 2016 LinkedInBetsy Bolte
 
Music videos and music genres
Music videos and music genresMusic videos and music genres
Music videos and music genresTubbsterinator
 

What's hot (20)

Questionnaire analysis
Questionnaire analysisQuestionnaire analysis
Questionnaire analysis
 
Analysis
AnalysisAnalysis
Analysis
 
Questionnaire 1
Questionnaire 1Questionnaire 1
Questionnaire 1
 
Planning production
Planning productionPlanning production
Planning production
 
Publication plan
Publication planPublication plan
Publication plan
 
Music questionnaire2015
Music questionnaire2015Music questionnaire2015
Music questionnaire2015
 
1st Draft of Power Players Magazine Editor Troy Taylor
1st Draft of Power Players Magazine Editor Troy Taylor1st Draft of Power Players Magazine Editor Troy Taylor
1st Draft of Power Players Magazine Editor Troy Taylor
 
Music video questionnaire
Music video questionnaireMusic video questionnaire
Music video questionnaire
 
Analysis of survey results
Analysis of survey resultsAnalysis of survey results
Analysis of survey results
 
Questionnaire analysis for a2
Questionnaire analysis for a2Questionnaire analysis for a2
Questionnaire analysis for a2
 
Researcb audience
Researcb audienceResearcb audience
Researcb audience
 
'Lyric' magazine
'Lyric' magazine'Lyric' magazine
'Lyric' magazine
 
The black keys
The black keysThe black keys
The black keys
 
Contents ideas
Contents ideasContents ideas
Contents ideas
 
Interviewee powerpoint
Interviewee powerpointInterviewee powerpoint
Interviewee powerpoint
 
Target audience analysis
Target audience analysisTarget audience analysis
Target audience analysis
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire results
 
Target Audience Survey
Target Audience SurveyTarget Audience Survey
Target Audience Survey
 
Betsy Bolte Artists & Labels 2016 LinkedIn
Betsy Bolte Artists & Labels 2016 LinkedInBetsy Bolte Artists & Labels 2016 LinkedIn
Betsy Bolte Artists & Labels 2016 LinkedIn
 
Music videos and music genres
Music videos and music genresMusic videos and music genres
Music videos and music genres
 

Viewers also liked

What's So Special About Mobile?
What's So Special About Mobile?What's So Special About Mobile?
What's So Special About Mobile?Sara Quinn
 
Getting Web Multi-Touch Working
Getting Web Multi-Touch Working Getting Web Multi-Touch Working
Getting Web Multi-Touch Working Aidan Wu
 
Mobile Marketing: How Consumers Use Mobile to Connect with Your Business
Mobile Marketing: How Consumers Use Mobile to Connect with Your BusinessMobile Marketing: How Consumers Use Mobile to Connect with Your Business
Mobile Marketing: How Consumers Use Mobile to Connect with Your BusinessSIXTY
 
Digiday and Yahoo! Present: Mobile Daily Habits, Moving Marketers Forward
Digiday and  Yahoo! Present: Mobile Daily Habits, Moving Marketers ForwardDigiday and  Yahoo! Present: Mobile Daily Habits, Moving Marketers Forward
Digiday and Yahoo! Present: Mobile Daily Habits, Moving Marketers ForwardDigiday
 
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...Form Factor is the Message: How the Devices We Use Shape the Messages and Con...
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...Future Insights
 
Android et ses petits secrets - SFR Régie - Mars 2012
Android et ses petits secrets  - SFR Régie - Mars 2012Android et ses petits secrets  - SFR Régie - Mars 2012
Android et ses petits secrets - SFR Régie - Mars 2012Romain Fonnier
 
Mobile forms - 10 DOs and DONTs
Mobile forms - 10 DOs and DONTsMobile forms - 10 DOs and DONTs
Mobile forms - 10 DOs and DONTsTomer Rosenthal
 
Sneak Peek at Google I/O 2014 Highlights
Sneak Peek at Google I/O 2014 HighlightsSneak Peek at Google I/O 2014 Highlights
Sneak Peek at Google I/O 2014 HighlightsTechAhead
 
On your mark, get set, mobile
On your mark, get set, mobileOn your mark, get set, mobile
On your mark, get set, mobileTiffany Beker
 
Placecast - E-commerce Paris 2012
Placecast - E-commerce Paris 2012Placecast - E-commerce Paris 2012
Placecast - E-commerce Paris 2012Petit Web
 
[refreshaustin] Adaptive Images in Responsive Web Design
[refreshaustin] Adaptive Images in Responsive Web Design[refreshaustin] Adaptive Images in Responsive Web Design
[refreshaustin] Adaptive Images in Responsive Web DesignChristopher Schmitt
 
Smart contents in multi screen
Smart contents in multi screenSmart contents in multi screen
Smart contents in multi screenSeungyul Kim
 
Mobile Marketing Trend Report - AUG 2012
Mobile Marketing Trend Report - AUG 2012Mobile Marketing Trend Report - AUG 2012
Mobile Marketing Trend Report - AUG 2012Daniel Wood
 
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entreprise
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entrepriseMobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entreprise
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entrepriseservicesmobiles.fr
 
Mobile Marketing Attitude 2013 - SNCD
Mobile Marketing Attitude 2013 - SNCDMobile Marketing Attitude 2013 - SNCD
Mobile Marketing Attitude 2013 - SNCDRomain Fonnier
 
10 Important Design Changes to iOS 7
10 Important Design Changes to iOS 710 Important Design Changes to iOS 7
10 Important Design Changes to iOS 7Ratio
 
Classement de la fréquentation des applications mobile OJD - Avril 2013
Classement de la fréquentation des applications mobile OJD - Avril 2013Classement de la fréquentation des applications mobile OJD - Avril 2013
Classement de la fréquentation des applications mobile OJD - Avril 2013Romain Fonnier
 

Viewers also liked (20)

What's So Special About Mobile?
What's So Special About Mobile?What's So Special About Mobile?
What's So Special About Mobile?
 
Getting Web Multi-Touch Working
Getting Web Multi-Touch Working Getting Web Multi-Touch Working
Getting Web Multi-Touch Working
 
Mobile Marketing: How Consumers Use Mobile to Connect with Your Business
Mobile Marketing: How Consumers Use Mobile to Connect with Your BusinessMobile Marketing: How Consumers Use Mobile to Connect with Your Business
Mobile Marketing: How Consumers Use Mobile to Connect with Your Business
 
Create great UIs for budget phones
Create great UIs for budget phonesCreate great UIs for budget phones
Create great UIs for budget phones
 
Digiday and Yahoo! Present: Mobile Daily Habits, Moving Marketers Forward
Digiday and  Yahoo! Present: Mobile Daily Habits, Moving Marketers ForwardDigiday and  Yahoo! Present: Mobile Daily Habits, Moving Marketers Forward
Digiday and Yahoo! Present: Mobile Daily Habits, Moving Marketers Forward
 
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...Form Factor is the Message: How the Devices We Use Shape the Messages and Con...
Form Factor is the Message: How the Devices We Use Shape the Messages and Con...
 
Chiffres du mobile en france
Chiffres du mobile en franceChiffres du mobile en france
Chiffres du mobile en france
 
Android et ses petits secrets - SFR Régie - Mars 2012
Android et ses petits secrets  - SFR Régie - Mars 2012Android et ses petits secrets  - SFR Régie - Mars 2012
Android et ses petits secrets - SFR Régie - Mars 2012
 
Mobile forms - 10 DOs and DONTs
Mobile forms - 10 DOs and DONTsMobile forms - 10 DOs and DONTs
Mobile forms - 10 DOs and DONTs
 
Sneak Peek at Google I/O 2014 Highlights
Sneak Peek at Google I/O 2014 HighlightsSneak Peek at Google I/O 2014 Highlights
Sneak Peek at Google I/O 2014 Highlights
 
On your mark, get set, mobile
On your mark, get set, mobileOn your mark, get set, mobile
On your mark, get set, mobile
 
Placecast - E-commerce Paris 2012
Placecast - E-commerce Paris 2012Placecast - E-commerce Paris 2012
Placecast - E-commerce Paris 2012
 
iPads in Elementary School
iPads in Elementary SchooliPads in Elementary School
iPads in Elementary School
 
[refreshaustin] Adaptive Images in Responsive Web Design
[refreshaustin] Adaptive Images in Responsive Web Design[refreshaustin] Adaptive Images in Responsive Web Design
[refreshaustin] Adaptive Images in Responsive Web Design
 
Smart contents in multi screen
Smart contents in multi screenSmart contents in multi screen
Smart contents in multi screen
 
Mobile Marketing Trend Report - AUG 2012
Mobile Marketing Trend Report - AUG 2012Mobile Marketing Trend Report - AUG 2012
Mobile Marketing Trend Report - AUG 2012
 
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entreprise
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entrepriseMobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entreprise
Mobile is not mobile - 7 phénomènes disruptifs* à prévoir dans l’entreprise
 
Mobile Marketing Attitude 2013 - SNCD
Mobile Marketing Attitude 2013 - SNCDMobile Marketing Attitude 2013 - SNCD
Mobile Marketing Attitude 2013 - SNCD
 
10 Important Design Changes to iOS 7
10 Important Design Changes to iOS 710 Important Design Changes to iOS 7
10 Important Design Changes to iOS 7
 
Classement de la fréquentation des applications mobile OJD - Avril 2013
Classement de la fréquentation des applications mobile OJD - Avril 2013Classement de la fréquentation des applications mobile OJD - Avril 2013
Classement de la fréquentation des applications mobile OJD - Avril 2013
 

Similar to I've got 10 million songs in my pocket. Now what?

Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire resultsnehajasmine
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire resultsnehajasmine
 
Results & analysis - Website,Cd,Music Video
Results & analysis - Website,Cd,Music VideoResults & analysis - Website,Cd,Music Video
Results & analysis - Website,Cd,Music Video224897
 
Materials Of Teaching (Music)
Materials Of Teaching (Music)Materials Of Teaching (Music)
Materials Of Teaching (Music)Iin Hermiyanto
 
Audience research results
Audience research resultsAudience research results
Audience research resultsPickwicken
 
Evaluation question 1
Evaluation question 1Evaluation question 1
Evaluation question 1Connor Ball
 
Audience Research - Part 2
Audience Research - Part 2 Audience Research - Part 2
Audience Research - Part 2 BHollis95
 
Audience Research - Part 2
Audience Research - Part 2 Audience Research - Part 2
Audience Research - Part 2 BHollis95
 
Audience Research Part 2
Audience Research Part 2 Audience Research Part 2
Audience Research Part 2 BHollis95
 
2. Research (FMP).pptx
2. Research (FMP).pptx2. Research (FMP).pptx
2. Research (FMP).pptxArchieDurham
 
The New Music Business Model- Stjames
The New Music Business Model- StjamesThe New Music Business Model- Stjames
The New Music Business Model- Stjamesmichaelstjames
 
The Seven Steps to Kickstarting a Successful Podcast
The Seven Steps to Kickstarting a Successful PodcastThe Seven Steps to Kickstarting a Successful Podcast
The Seven Steps to Kickstarting a Successful PodcastNapierPR
 
Questionnaire Analysis - Media Studies A2
Questionnaire Analysis - Media Studies A2Questionnaire Analysis - Media Studies A2
Questionnaire Analysis - Media Studies A2MissKylieLee
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire resultsShakiraBBS
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire resultsthatasiankim
 
Songwriting Studies Week 1, introduction
Songwriting Studies Week 1, introductionSongwriting Studies Week 1, introduction
Songwriting Studies Week 1, introductionMegan Berry
 

Similar to I've got 10 million songs in my pocket. Now what? (20)

Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire results
 
a
aa
a
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire results
 
Results & analysis - Website,Cd,Music Video
Results & analysis - Website,Cd,Music VideoResults & analysis - Website,Cd,Music Video
Results & analysis - Website,Cd,Music Video
 
Evaluation Q3
Evaluation Q3Evaluation Q3
Evaluation Q3
 
Materials Of Teaching (Music)
Materials Of Teaching (Music)Materials Of Teaching (Music)
Materials Of Teaching (Music)
 
Audience research results
Audience research resultsAudience research results
Audience research results
 
Evaluation question 1
Evaluation question 1Evaluation question 1
Evaluation question 1
 
Audience Research - Part 2
Audience Research - Part 2 Audience Research - Part 2
Audience Research - Part 2
 
Audience Research - Part 2
Audience Research - Part 2 Audience Research - Part 2
Audience Research - Part 2
 
Audience Research Part 2
Audience Research Part 2 Audience Research Part 2
Audience Research Part 2
 
2. Research (FMP).pptx
2. Research (FMP).pptx2. Research (FMP).pptx
2. Research (FMP).pptx
 
The New Music Business Model- Stjames
The New Music Business Model- StjamesThe New Music Business Model- Stjames
The New Music Business Model- Stjames
 
The Seven Steps to Kickstarting a Successful Podcast
The Seven Steps to Kickstarting a Successful PodcastThe Seven Steps to Kickstarting a Successful Podcast
The Seven Steps to Kickstarting a Successful Podcast
 
Questionnaire Analysis - Media Studies A2
Questionnaire Analysis - Media Studies A2Questionnaire Analysis - Media Studies A2
Questionnaire Analysis - Media Studies A2
 
Interview 2
Interview  2Interview  2
Interview 2
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire results
 
Questionnaire
QuestionnaireQuestionnaire
Questionnaire
 
Questionnaire results
Questionnaire resultsQuestionnaire results
Questionnaire results
 
Songwriting Studies Week 1, introduction
Songwriting Studies Week 1, introductionSongwriting Studies Week 1, introduction
Songwriting Studies Week 1, introduction
 

More from Paul Lamere

Sxsw 2015 - How we listen to music
Sxsw 2015 - How we listen to musicSxsw 2015 - How we listen to music
Sxsw 2015 - How we listen to musicPaul Lamere
 
Echo nest-api-boston-2012
Echo nest-api-boston-2012Echo nest-api-boston-2012
Echo nest-api-boston-2012Paul Lamere
 
Data Mining Music
Data Mining MusicData Mining Music
Data Mining MusicPaul Lamere
 
Finding Music With Pictures: Using Visualization for Discovery
Finding Music With Pictures: Using Visualization for DiscoveryFinding Music With Pictures: Using Visualization for Discovery
Finding Music With Pictures: Using Visualization for DiscoveryPaul Lamere
 
The Echo Nest workshop for Boston Music Hack Day
The Echo Nest workshop for Boston Music Hack DayThe Echo Nest workshop for Boston Music Hack Day
The Echo Nest workshop for Boston Music Hack DayPaul Lamere
 
Using Visualizations for Music Discovery
Using Visualizations for Music DiscoveryUsing Visualizations for Music Discovery
Using Visualizations for Music DiscoveryPaul Lamere
 
Help! My iPod thinks I'm emo.
Help! My iPod thinks I'm emo.Help! My iPod thinks I'm emo.
Help! My iPod thinks I'm emo.Paul Lamere
 
Social Tags and Music Information Retrieval (Part I)
Social Tags and Music Information Retrieval (Part I)Social Tags and Music Information Retrieval (Part I)
Social Tags and Music Information Retrieval (Part I)Paul Lamere
 
Social Tags and Music Information Retrieval (Part II)
Social Tags and Music Information Retrieval (Part II)Social Tags and Music Information Retrieval (Part II)
Social Tags and Music Information Retrieval (Part II)Paul Lamere
 

More from Paul Lamere (9)

Sxsw 2015 - How we listen to music
Sxsw 2015 - How we listen to musicSxsw 2015 - How we listen to music
Sxsw 2015 - How we listen to music
 
Echo nest-api-boston-2012
Echo nest-api-boston-2012Echo nest-api-boston-2012
Echo nest-api-boston-2012
 
Data Mining Music
Data Mining MusicData Mining Music
Data Mining Music
 
Finding Music With Pictures: Using Visualization for Discovery
Finding Music With Pictures: Using Visualization for DiscoveryFinding Music With Pictures: Using Visualization for Discovery
Finding Music With Pictures: Using Visualization for Discovery
 
The Echo Nest workshop for Boston Music Hack Day
The Echo Nest workshop for Boston Music Hack DayThe Echo Nest workshop for Boston Music Hack Day
The Echo Nest workshop for Boston Music Hack Day
 
Using Visualizations for Music Discovery
Using Visualizations for Music DiscoveryUsing Visualizations for Music Discovery
Using Visualizations for Music Discovery
 
Help! My iPod thinks I'm emo.
Help! My iPod thinks I'm emo.Help! My iPod thinks I'm emo.
Help! My iPod thinks I'm emo.
 
Social Tags and Music Information Retrieval (Part I)
Social Tags and Music Information Retrieval (Part I)Social Tags and Music Information Retrieval (Part I)
Social Tags and Music Information Retrieval (Part I)
 
Social Tags and Music Information Retrieval (Part II)
Social Tags and Music Information Retrieval (Part II)Social Tags and Music Information Retrieval (Part II)
Social Tags and Music Information Retrieval (Part II)
 

Recently uploaded

Vertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsVertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsMiki Katsuragi
 
Dev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebDev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebUiPathCommunity
 
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxSAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxNavinnSomaal
 
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Mark Simos
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clashcharlottematthew16
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupFlorian Wilhelm
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Manik S Magar
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Mattias Andersson
 
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level pieceStory boards and shot lists for my a level piece
Story boards and shot lists for my a level piececharlottematthew16
 
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek SchlawackFwdays
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr BaganFwdays
 
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii SoldatenkoFwdays
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Patryk Bandurski
 
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024BookNet Canada
 
DevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenDevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenHervé Boutemy
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024The Digital Insurer
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticscarlostorres15106
 

Recently uploaded (20)

Vertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsVertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering Tips
 
Dev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio WebDev Dives: Streamline document processing with UiPath Studio Web
Dev Dives: Streamline document processing with UiPath Studio Web
 
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxSAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptx
 
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clash
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special EditionDMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
 
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?
 
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level pieceStory boards and shot lists for my a level piece
Story boards and shot lists for my a level piece
 
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
 
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan
 
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
 
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
 
DevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache MavenDevoxxFR 2024 Reproducible Builds with Apache Maven
DevoxxFR 2024 Reproducible Builds with Apache Maven
 
My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024My INSURER PTE LTD - Insurtech Innovation Award 2024
My INSURER PTE LTD - Insurtech Innovation Award 2024
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
 

I've got 10 million songs in my pocket. Now what?

  • 1. I have 10 million songs in my pocket. Now what? Paul Lamere ACM Recommender Systems 2012 Photo (CC) by Jason Rogers
  • 2. The challenge of music recommendation A recommendation that no human would make If you like Britney Spears ... You might like the Report on Pre-War Intelligence
  • 3. The challenge of music recommendation WTF - Why the Freakomendations? Why do we see such bad music recommendations? Music is Special* *of course every domain is special
  • 4. The challenge of music recommendation Why do we see such bad music recommendations? Understanding the domain is critical to the success of a recommender
  • 5. The challenge of music recommendation Why do we see such bad music recommendations? 10 things to consider when building a music recommender
  • 6. #1 - Very large item space 10 things to consider when building a music recommender
  • 7. #2 - Very low cost per item 10 things to consider when building a music recommender
  • 8. #2 - Very low cost per item 10 things to consider when building a music recommender
  • 9. #3 - Low consumption time 10 things to consider when building a music recommender The average song length is around 4 minutes E Book Reader: 24 books a year Music subscriber: 25 songs a day Pew Internet Project.
  • 10. #4 - Highly Interactive 10 things to consider when building a music recommender A typical music recommender ... ... in 1999, but not anymore
  • 11. #4 - Highly Interactive 10 things to consider when building a music recommender Today’s recommender Recommendations are integrated into the listening experience
  • 12. #4 - Highly Interactive 10 things to consider when building a music recommender Today’s recommender Minimal explicit feedback
  • 13. #4 - Highly Interactive 10 things to consider when building a music recommender Today’s recommender Much Implicit Feedback - Playing - Skipping - Repeating - Adjusting the volume - Sharing with friends - Adding to a playlist - Repeating the song - Inspecting song info
  • 14. #5 - Very high per-item reuse 10 things to consider when building a music recommender songs[“as time goes by”].playcount += 1
  • 15. #6 - Large personal collections 10 things to consider when building a music recommender
  • 16. Personal Music Discovery Challenge There’s a long tail in my iPod Listener Study Listeners 5,000 Average Songs 3,500 Per User Percent of songs 65% never listened to
  • 17. #7 - Consumed in sequences 10 things to consider when building a music recommender A good playlist is a balance of: • Coherence • Familiarity • Discovery • Variety • Surprise ... in a pleasing order
  • 18. #8 - Highly contextual usage 10 things to consider when building a music recommender
  • 19. #9 - Highly passionate users 10 things to consider when building a music recommender
  • 20. Let’s pause for a quiz Why is this formula troublesome for music recommendation and discovery? (ΔMī¹=αΣDi[n][ΣFij[n-1]+Fexti[[n̄¹]])
  • 21. Let’s pause for a quiz Why is this formula troublesome for music recommendation and discovery? Because it is the name of a song by Aphex Twin (ΔMī¹=αΣDi[n][ΣFij[n-1]+Fexti[[n̄¹]])
  • 22. #10 - OMG Metadata 10 things to consider when building a music recommender The The Duran Duran Duran !!! ††† ///▲▲▲ ▼□■□■□■ Various Artists
  • 23. #10 - OMG Metadata 10 things to consider when building a music recommender
  • 24. Music Taste is Irrational The challenge Music Music I Like You Like Music I Used To Like Get this t-shirt at dieselsweeties.com
  • 25. I have 10 million songs in my pocket. Now what? Paul Lamere Paul@echonest.com MusicMachinery.com @plamere Photo (CC) by Jason Rogers