SlideShare a Scribd company logo
1 of 19
Download to read offline
Users and Noise: The Magic Barrier of Recommender Systems



 Alan Said, Brijnesh J. Jain, Sascha Narr, Till Plumbaum
  Competence Center Information Retrieval & Machine Learning


 @alansaid, @saschanarr, @matip
Outline

► The Magic Barrier
► Empirical Risk Minimization

► Deriving the Magic Barrier

► User Study

► Conclusion




           20 July 2012   The Magic Barrier   2
The Magic Barrier




         20 July 2012   The Magic Barrier   3
The Magic Barrier

► No magic involved....
► Coined by Herlocker et al. in 2004

      “...an algorithm cannot be more accurate than the variance in
       a user’s ratings for the same item.”
      The maximum level of prediction that a recommender
       algorithm can attain.



►   What does this mean?




            20 July 2012   The Magic Barrier                       4
The Magic Barrier




         20 July 2012   The Magic Barrier   5
The Magic Barrier

►   Even a “perfect” recommender should not reach RMSE = 0 or
    Precision @ N = 1

►   Why?
       People are inconsistent and noisy in their ratings
       “perfect” accuracy is not perfect

►   So?
       Knowing the highest possible level of accuracy, we can stop
        optimizing our algorithms at “perfect” (before overfitting)




              20 July 2012    The Magic Barrier                       6
The Magic Barrier




So – how do we find the magic barrier?

We employ the Empirical Risk Minimization principle and a
 statistical model for user inconsistencies




           20 July 2012   The Magic Barrier                 7
The Magic Barrier – User Inconsistencies

Assumption:
    If a user were to re-rate all previously rated items, keeping in
     mind the inconsistency, the ratings would differ, i.e.
            𝑟 𝑢𝑖 = 𝜇 𝑢𝑖 + 𝜀 𝑢𝑖

        where
           𝜇 𝑢𝑖 is the expected rating, and
           𝜀 𝑢𝑖 the rating error (has zero mean)




            20 July 2012          The Magic Barrier                     8
Empirical Risk Minimization

►   … is a principle in statistical learning theory which defines a
    family of learning algorithms and is used to give theoretical
    bounds on the performance of learning
    algorithms.[Wikipedia]




              20 July 2012   The Magic Barrier                        9
Empirical Risk Minimization

►   We formulate our risk function as
       𝑅 𝑓 = 𝑢,𝑖,𝑟 𝑝 𝑢, 𝑖, 𝑟 𝑓 𝑢, 𝑖 − 𝑟 2                     The prediction error

        The probability of user u rating item i with score r

►   Keeping the assumption in mind, we formulate the risk for a
    true, unknown, rating function as the sum of the noise
    variance, i.e.
        𝑅 𝑓∗ = 𝑢,𝑖 𝑝 𝑢, 𝑖 𝕍 𝜀 𝑢𝑖
           where 𝕍 𝜀 𝑢𝑖 is the noise variance




               20 July 2012         The Magic Barrier                                 10
Deriving the Magic Barrier

►   We want to express the risk function in terms of a magic barrier
    for RMSE – we take the root of the risk function

       ℬ 𝒰×ℐ =              𝑢,𝑖   𝑝 𝑢, 𝑖 𝕍 𝜀 𝑢𝑖

       RMSE=0 iff 𝜀 𝑢𝑖 = 0 over all ratings users and items
► In terms of RMSE we can express this as

       𝐸 𝑅𝑀𝑆𝐸 𝑓 = ℬ 𝒰×ℐ + 𝐸 𝑓 > ℬ 𝒰×ℐ
       where 𝐸 𝑓 is the error




              20 July 2012             The Magic Barrier          11
Estimating the Magic Barrier

1.   For each user-item pair in our population
      a) Sample ratings on a regular basis, i.e. re-ratings
      b) Estimate the expected value of ratings
                                                     𝑚
                                                1
                                     𝜇 𝑢𝑖     =           𝑟 𝑡 𝑢𝑖
                                                𝑚
                                                    𝑡=1

     c. Estimate the rating variance
                                          𝑚
                                     1                                  2
                        𝜀 𝑢𝑖   2
                                   =
                                     𝑚
                                               𝜇 𝑢𝑖 −        𝑟𝑡    𝑢𝑖
                                         𝑡=1

2.   Estimate the magic barrier by taking the average
                                                1
                                    ℬ=                                  𝜀 𝑢𝑖 2
                                                𝒳
                                                          𝑢𝑖 ∈𝒳




             20 July 2012                The Magic Barrier                       12
A real-world user study




     20 July 2012   The Magic Barrier   13
A User Study

► We teamed up with moviepilot.de
      Germany’s largest online movie recommendation community
      Ratings scale 1-10 stars (Netflix: 1-5 stars)
► Created a re-rating UI

      Users were asked to re-rate at least 20 movies
        1 new rating (so-called opinions) per movie
     Collected data:
        306 users
        6,299 new opinions
        2,329 movies




           20 July 2012       The Magic Barrier              14
A User Study




      User study                             moviepilot




          20 July 2012   The Magic Barrier                15
A User Study


                    ~4 ratings steps          Room for improvement

                                        ~1 rating steps




 Predictions vs               Ratings above                    Ratings below
    Ratings                   user’s average                   user’s average
                    Overall                   Opinions above                Opinions below
                  Magic Barrier               user’s average                user’s average

              20 July 2012         The Magic Barrier                                  16
Conclusion

► We created a mathematical characterization of the magic
  barrier
► We performed a user study on a commercial movie

  recommendation website and estimated its magic barrier
► We concluded the commercial recommender engine still has

  room for improvement

►   No magic




               20 July 2012   The Magic Barrier              17
More?

►   Estimating the Magic Barrier of Recommender Systems: A User Study
         SIGIR 2012

►   Magic Barrier explained
       http://irml.dailab.de

►   Movie rating and explanation user study
       http://j.mp/ratingexplain

►   Recommender Systems Wiki
        www.recsyswiki.com

►   Recommender Systems Challenge
        www.recsyschallenge.com


               20 July 2012     The Magic Barrier                       18
Questions?




►   Thank You for Listening!




              20 July 2012     The Magic Barrier   19

More Related Content

More from Alan Said

Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...
Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...
Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...Alan Said
 
The Magic Barrier of Recommender Systems - No Magic, Just Ratings
The Magic Barrier of Recommender Systems - No Magic, Just RatingsThe Magic Barrier of Recommender Systems - No Magic, Just Ratings
The Magic Barrier of Recommender Systems - No Magic, Just RatingsAlan Said
 
A Top-N Recommender System Evaluation Protocol Inspired by Deployed Systems
A Top-N Recommender System Evaluation Protocol Inspired by Deployed SystemsA Top-N Recommender System Evaluation Protocol Inspired by Deployed Systems
A Top-N Recommender System Evaluation Protocol Inspired by Deployed SystemsAlan Said
 
Information Retrieval and User-centric Recommender System Evaluation
Information Retrieval and User-centric Recommender System EvaluationInformation Retrieval and User-centric Recommender System Evaluation
Information Retrieval and User-centric Recommender System EvaluationAlan Said
 
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...Alan Said
 
A 3D Approach to Recommender System Evaluation
A 3D Approach to Recommender System EvaluationA 3D Approach to Recommender System Evaluation
A 3D Approach to Recommender System EvaluationAlan Said
 
State of RecSys: Recap of RecSys 2012
State of RecSys: Recap of RecSys 2012State of RecSys: Recap of RecSys 2012
State of RecSys: Recap of RecSys 2012Alan Said
 
RecSysChallenge Opening
RecSysChallenge OpeningRecSysChallenge Opening
RecSysChallenge OpeningAlan Said
 
Best Practices in Recommender System Challenges
Best Practices in Recommender System ChallengesBest Practices in Recommender System Challenges
Best Practices in Recommender System ChallengesAlan Said
 
Estimating the Magic Barrier of Recommender Systems: A User Study
Estimating the Magic Barrier of Recommender Systems: A User StudyEstimating the Magic Barrier of Recommender Systems: A User Study
Estimating the Magic Barrier of Recommender Systems: A User StudyAlan Said
 
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...Alan Said
 
CaRR 2012 Opening Presentation
CaRR 2012 Opening PresentationCaRR 2012 Opening Presentation
CaRR 2012 Opening PresentationAlan Said
 
Personalizing Tags: A Folksonomy-like Approach for Recommending Movies
Personalizing Tags: A Folksonomy-like Approach for Recommending MoviesPersonalizing Tags: A Folksonomy-like Approach for Recommending Movies
Personalizing Tags: A Folksonomy-like Approach for Recommending MoviesAlan Said
 
Inferring Contextual User Profiles - Improving Recommender Performance
Inferring Contextual User Profiles - Improving Recommender PerformanceInferring Contextual User Profiles - Improving Recommender Performance
Inferring Contextual User Profiles - Improving Recommender PerformanceAlan Said
 
Using Social- and Pseudo-Social Networks to Improve Recommendation Quality
Using Social- and Pseudo-Social Networks to Improve Recommendation QualityUsing Social- and Pseudo-Social Networks to Improve Recommendation Quality
Using Social- and Pseudo-Social Networks to Improve Recommendation QualityAlan Said
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender SystemsAlan Said
 

More from Alan Said (16)

Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...
Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...
Comparative Recommender System Evaluation: Benchmarking Recommendation Frame...
 
The Magic Barrier of Recommender Systems - No Magic, Just Ratings
The Magic Barrier of Recommender Systems - No Magic, Just RatingsThe Magic Barrier of Recommender Systems - No Magic, Just Ratings
The Magic Barrier of Recommender Systems - No Magic, Just Ratings
 
A Top-N Recommender System Evaluation Protocol Inspired by Deployed Systems
A Top-N Recommender System Evaluation Protocol Inspired by Deployed SystemsA Top-N Recommender System Evaluation Protocol Inspired by Deployed Systems
A Top-N Recommender System Evaluation Protocol Inspired by Deployed Systems
 
Information Retrieval and User-centric Recommender System Evaluation
Information Retrieval and User-centric Recommender System EvaluationInformation Retrieval and User-centric Recommender System Evaluation
Information Retrieval and User-centric Recommender System Evaluation
 
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...
User-Centric Evaluation of a K-Furthest Neighbor Collaborative Filtering Reco...
 
A 3D Approach to Recommender System Evaluation
A 3D Approach to Recommender System EvaluationA 3D Approach to Recommender System Evaluation
A 3D Approach to Recommender System Evaluation
 
State of RecSys: Recap of RecSys 2012
State of RecSys: Recap of RecSys 2012State of RecSys: Recap of RecSys 2012
State of RecSys: Recap of RecSys 2012
 
RecSysChallenge Opening
RecSysChallenge OpeningRecSysChallenge Opening
RecSysChallenge Opening
 
Best Practices in Recommender System Challenges
Best Practices in Recommender System ChallengesBest Practices in Recommender System Challenges
Best Practices in Recommender System Challenges
 
Estimating the Magic Barrier of Recommender Systems: A User Study
Estimating the Magic Barrier of Recommender Systems: A User StudyEstimating the Magic Barrier of Recommender Systems: A User Study
Estimating the Magic Barrier of Recommender Systems: A User Study
 
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...
Analyzing Weighting Schemes in Collaborative Filtering: Cold Start, Post Cold...
 
CaRR 2012 Opening Presentation
CaRR 2012 Opening PresentationCaRR 2012 Opening Presentation
CaRR 2012 Opening Presentation
 
Personalizing Tags: A Folksonomy-like Approach for Recommending Movies
Personalizing Tags: A Folksonomy-like Approach for Recommending MoviesPersonalizing Tags: A Folksonomy-like Approach for Recommending Movies
Personalizing Tags: A Folksonomy-like Approach for Recommending Movies
 
Inferring Contextual User Profiles - Improving Recommender Performance
Inferring Contextual User Profiles - Improving Recommender PerformanceInferring Contextual User Profiles - Improving Recommender Performance
Inferring Contextual User Profiles - Improving Recommender Performance
 
Using Social- and Pseudo-Social Networks to Improve Recommendation Quality
Using Social- and Pseudo-Social Networks to Improve Recommendation QualityUsing Social- and Pseudo-Social Networks to Improve Recommendation Quality
Using Social- and Pseudo-Social Networks to Improve Recommendation Quality
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 

Recently uploaded

Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Enterprise Knowledge
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024The Digital Insurer
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
HTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation StrategiesHTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation StrategiesBoston Institute of Analytics
 
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUnderstanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUK Journal
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoffsammart93
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Scriptwesley chun
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘RTylerCroy
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Drew Madelung
 
Tech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfTech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfhans926745
 

Recently uploaded (20)

Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
 
HTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation StrategiesHTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation Strategies
 
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUnderstanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
Tech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfTech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdf
 

Users and Noise: The Magic Barrier of Recommender Systems

  • 1. Users and Noise: The Magic Barrier of Recommender Systems Alan Said, Brijnesh J. Jain, Sascha Narr, Till Plumbaum Competence Center Information Retrieval & Machine Learning @alansaid, @saschanarr, @matip
  • 2. Outline ► The Magic Barrier ► Empirical Risk Minimization ► Deriving the Magic Barrier ► User Study ► Conclusion 20 July 2012 The Magic Barrier 2
  • 3. The Magic Barrier 20 July 2012 The Magic Barrier 3
  • 4. The Magic Barrier ► No magic involved.... ► Coined by Herlocker et al. in 2004  “...an algorithm cannot be more accurate than the variance in a user’s ratings for the same item.”  The maximum level of prediction that a recommender algorithm can attain. ► What does this mean? 20 July 2012 The Magic Barrier 4
  • 5. The Magic Barrier 20 July 2012 The Magic Barrier 5
  • 6. The Magic Barrier ► Even a “perfect” recommender should not reach RMSE = 0 or Precision @ N = 1 ► Why?  People are inconsistent and noisy in their ratings  “perfect” accuracy is not perfect ► So?  Knowing the highest possible level of accuracy, we can stop optimizing our algorithms at “perfect” (before overfitting) 20 July 2012 The Magic Barrier 6
  • 7. The Magic Barrier So – how do we find the magic barrier? We employ the Empirical Risk Minimization principle and a statistical model for user inconsistencies 20 July 2012 The Magic Barrier 7
  • 8. The Magic Barrier – User Inconsistencies Assumption:  If a user were to re-rate all previously rated items, keeping in mind the inconsistency, the ratings would differ, i.e. 𝑟 𝑢𝑖 = 𝜇 𝑢𝑖 + 𝜀 𝑢𝑖  where  𝜇 𝑢𝑖 is the expected rating, and  𝜀 𝑢𝑖 the rating error (has zero mean) 20 July 2012 The Magic Barrier 8
  • 9. Empirical Risk Minimization ► … is a principle in statistical learning theory which defines a family of learning algorithms and is used to give theoretical bounds on the performance of learning algorithms.[Wikipedia] 20 July 2012 The Magic Barrier 9
  • 10. Empirical Risk Minimization ► We formulate our risk function as  𝑅 𝑓 = 𝑢,𝑖,𝑟 𝑝 𝑢, 𝑖, 𝑟 𝑓 𝑢, 𝑖 − 𝑟 2 The prediction error The probability of user u rating item i with score r ► Keeping the assumption in mind, we formulate the risk for a true, unknown, rating function as the sum of the noise variance, i.e.  𝑅 𝑓∗ = 𝑢,𝑖 𝑝 𝑢, 𝑖 𝕍 𝜀 𝑢𝑖  where 𝕍 𝜀 𝑢𝑖 is the noise variance 20 July 2012 The Magic Barrier 10
  • 11. Deriving the Magic Barrier ► We want to express the risk function in terms of a magic barrier for RMSE – we take the root of the risk function  ℬ 𝒰×ℐ = 𝑢,𝑖 𝑝 𝑢, 𝑖 𝕍 𝜀 𝑢𝑖  RMSE=0 iff 𝜀 𝑢𝑖 = 0 over all ratings users and items ► In terms of RMSE we can express this as  𝐸 𝑅𝑀𝑆𝐸 𝑓 = ℬ 𝒰×ℐ + 𝐸 𝑓 > ℬ 𝒰×ℐ  where 𝐸 𝑓 is the error 20 July 2012 The Magic Barrier 11
  • 12. Estimating the Magic Barrier 1. For each user-item pair in our population a) Sample ratings on a regular basis, i.e. re-ratings b) Estimate the expected value of ratings 𝑚 1 𝜇 𝑢𝑖 = 𝑟 𝑡 𝑢𝑖 𝑚 𝑡=1 c. Estimate the rating variance 𝑚 1 2 𝜀 𝑢𝑖 2 = 𝑚 𝜇 𝑢𝑖 − 𝑟𝑡 𝑢𝑖 𝑡=1 2. Estimate the magic barrier by taking the average 1 ℬ= 𝜀 𝑢𝑖 2 𝒳 𝑢𝑖 ∈𝒳 20 July 2012 The Magic Barrier 12
  • 13. A real-world user study 20 July 2012 The Magic Barrier 13
  • 14. A User Study ► We teamed up with moviepilot.de  Germany’s largest online movie recommendation community  Ratings scale 1-10 stars (Netflix: 1-5 stars) ► Created a re-rating UI  Users were asked to re-rate at least 20 movies  1 new rating (so-called opinions) per movie  Collected data:  306 users  6,299 new opinions  2,329 movies 20 July 2012 The Magic Barrier 14
  • 15. A User Study User study moviepilot 20 July 2012 The Magic Barrier 15
  • 16. A User Study ~4 ratings steps Room for improvement ~1 rating steps Predictions vs Ratings above Ratings below Ratings user’s average user’s average Overall Opinions above Opinions below Magic Barrier user’s average user’s average 20 July 2012 The Magic Barrier 16
  • 17. Conclusion ► We created a mathematical characterization of the magic barrier ► We performed a user study on a commercial movie recommendation website and estimated its magic barrier ► We concluded the commercial recommender engine still has room for improvement ► No magic 20 July 2012 The Magic Barrier 17
  • 18. More? ► Estimating the Magic Barrier of Recommender Systems: A User Study  SIGIR 2012 ► Magic Barrier explained  http://irml.dailab.de ► Movie rating and explanation user study  http://j.mp/ratingexplain ► Recommender Systems Wiki  www.recsyswiki.com ► Recommender Systems Challenge  www.recsyschallenge.com 20 July 2012 The Magic Barrier 18
  • 19. Questions? ► Thank You for Listening! 20 July 2012 The Magic Barrier 19