[Final]collaborative filtering and recommender systems
- 1. Presentation for the Data Mining course Collaborative filtering and recommender systems Presented by: Falitokiniaina RABEARISON 30-10-2014 1
- 2. Collaborative filtering and Recommender Systems Life is too short! 2
- 3. Collaborative filtering and Recommender Systems Recommender systems 3
- 4. Collaborative filtering and Recommender Systems AGENDA Recommender systems Algorithms o Content based o Collaborative Filtering (User Based / Item Based) Challenges & Comparison 4
- 6. Collaborative filtering and Recommender Systems • RS seen as a function • Given: – User model (e g. . ratings ratings, preferences preferences, demographics demographics, situational situational context) context) – Items (with or without description of item characteristics). • Find: • - Relevance score. Used for ranking. • Finally: – Recommend items that are assumed to be relevant 6
- 7. Collaborative filtering and Recommender Systems RS > Paradigms of recommender systems 7
- 8. Collaborative filtering and Recommender Systems RS > Paradigms of recommender systems 8
- 9. Collaborative filtering and Recommender Systems RS > Paradigms of recommender systems 9
- 10. Collaborative filtering and Recommender Systems RS >Paradigms of recommender systems 10
- 11. Collaborative filtering and Recommender Systems RS > Paradigms of recommender systems 11
- 12. Collaborative filtering and Recommender Systems RS > Paradigms of recommender systems 12
- 13. Collaborative filtering and Recommender Systems RS > Results 13
- 14. Collaborative filtering and Recommender Systems Recommender approaches 14
- 15. ALGORITHMS CONTENT BASED FILTERING (CB) COLLABORATIVE FILTERING (CF) 15
- 16. Collaborative filtering and Recommender Systems CB > Content based algorithms • These rely on the implicit data on the domain « in a movie recommendation site, this could be the director information, movie length, PG rating, cast etc. » « For the song recommendation this could be song date, other albums/songs from the same group, type of the song (jazz, classi, rock, etc.) » • Implicit data is used in generating recommendations « You see that a user has rated high to Brad Pitt movies, so you recommend her Babel » 16
- 17. Collaborative filtering and Recommender Systems CB > OBJECT 17
- 18. Collaborative filtering and Recommender Systems CB > OBJECT INFORMATION 18
- 19. Collaborative filtering and Recommender Systems CB > FEATURE SET 19
- 20. 20
- 21. Collaborative filtering and Recommender Systems CB > SIMILARITY MATRIX 21
- 22. Collaborative filtering and Recommender Systems CB > SIMILARITY MEASURE 22
- 23. Collaborative filtering and Recommender Systems CB > SIMILARITY MEASURE 23
- 24. Collaborative filtering and Recommender Systems CB > SIMILARITY MATRIX 24
- 25. Collaborative filtering and Recommender Systems CB > SIMILARITY SORTING 25
- 26. Collaborative filtering and Recommender Systems CB > K-NEAREST NEIGHBOR (knn) 26
- 27. ALGORITHMS CONTENT BASED FILTERING (CB) COLLABORATIVE FILTERING (CF) 27
- 28. Collaborative filtering and Recommender Systems CF > Collaborative Filtering algorithms • Other users have impact on the recommendations, users generate recommendation implicitly. • Similar users to the active user (user that recommendations are prepared for) are found. • By weighting the users, a recommendation list is prepared from other user data. 28
- 29. Collaborative filtering and Recommender Systems CF > Basic idea
- 30. Collaborative filtering and Recommender Systems CF > Basic idea
- 31. Collaborative filtering and Recommender Systems CF > Collaborative Filtering Techniques
- 32. Collaborative filtering and Recommender Systems CF > USER & ITEM
- 33. Collaborative filtering and Recommender Systems CB > ORDER DATA
- 34. Collaborative filtering and Recommender Systems CF > ORDER DATA (cont.)
- 35. Collaborative filtering and Recommender Systems CF > ORDER DATA (cont.)
- 36. Collaborative filtering and Recommender Systems CF > VECTOR & DIMENSION
- 37. Collaborative filtering and Recommender Systems CF > VECTOR & DIMENSION
- 38. Collaborative filtering and Recommender Systems CF > VECTORS
- 39. Collaborative filtering and Recommender Systems CF > VECTORS
- 40. Collaborative filtering and Recommender Systems CF > SIMILARITY CALCULATION
- 41. Collaborative filtering and Recommender Systems CF > USER SIMILARITY MATRIX
- 42. Collaborative filtering and Recommender Systems CF > SIMILARITY CALCULATION
- 43. Collaborative filtering and Recommender Systems CF > SIMILARITY CALCULATION
- 44. Collaborative filtering and Recommender Systems CF > SIMILARITY CALCULATION EXAMPLE
- 45. Collaborative filtering and Recommender Systems CF > K-NEAREST-NEIGHBOR
- 46. Collaborative filtering and Recommender Systems CF > K-NEAREST-NEIGHBOR
- 47. Collaborative filtering and Recommender Systems CF > NEIGHBORS’ ORDER
- 48. Collaborative filtering and Recommender Systems CF > REMOVE BOUGHT ITEMS
- 49. Collaborative filtering and Recommender Systems CF > CALCULATING FINAL SCORE
- 50. Collaborative filtering and Recommender Systems CF > OTHER SIMILARITY MEASURES More at: http://favi.com.vn/wp-content/uploads/2012/05/pg049_Similarity_Measures_for_Text_Document_Clustering.pdf
- 51. Collaborative filtering and Recommender Systems CF > Collaborative Filtering Techniques
- 52. Collaborative filtering and Recommender Systems CF > ITEM SIMILARITY MATRIX
- 53. CHALLENGES AND COMPARISON 53
- 54. Collaborative filtering and Recommender Systems CHALLENGES • Dimensionality reduction (eg. Use PCA) • Input data sparsity • Overfitting to training data set 54
- 55. Collaborative filtering and Recommender Systems Advantages of CF over CF 55 Content based Recommender Collaborative based Recommender
- 56. Collaborative filtering and Recommender Systems 56
Editor's Notes
- #3: We don’t have time to watch all the movies, listen to all the music, read every book, etc … Which digital camera should I buy? What is the best holiday for me and my family? Which is the best investment for supporting the education of my children children? Which movie should I rent? Which web sites will I find interesting interesting? Which book should I buy for my next vacation? Which degree and university are the best for my future?
- #9: Collaborative Filtering Content‐based Filtering Knowledge‐Based Recommendations Hybridization Strategies
- #10: Collaborative Filtering Content‐based Filtering Knowledge‐Based Recommendations Hybridization Strategies
- #11: Collaborative Filtering Content‐based Filtering Knowledge‐Based Recommendations Hybridization Strategies
- #12: Collaborative Filtering Content‐based Filtering Knowledge‐Based Recommendations Hybridization Strategies
- #13: Collaborative Filtering Content‐based Filtering Knowledge‐Based Recommendations Hybridization Strategies
- #15: Complexity and accuracy
- #24: How to compute f(attribute)
- #28: User-user collaborative filtering