Recommender System Seminar

Transcript

1. RECOMMENDER SYSTEM
2. Building Recommender System in Industries
3. 1. Background: Definition, Problem Construction and Possible Solutions. 2. SVD-like for Recommender System: Matrix Factorization to model matrix completion problem 3. Implicit and Explicit signal: How to model Explicit and Implicit Signal with SVD-like RecSys 4. Recommender System workflow: Estimate, Filtering, Ranking and Randomization 5. Recommender System Cases Content
4. Background: Recommender System in Industries
5. • Chris Anderson in “The Long Tail ◦ “We are leaving the age of information and entering the age of recommendation” • CNN Money, “The race to create a 'smart' Google”: ◦ “The Web, they say, is leaving the era of search and entering one of discovery. What's the difference? Search is what you do when you're looking for something. Discovery is when something wonderful that you didn't know existed, or didn't know how to ask for, finds you.” Source: compiled from Xavier Amatriain MLSS slides (2014) Background
6. Background
7. Background
8. • Value of Recommendation ◦ Netflix: 2/3 of the movies watched are recommended ◦ Google News: recommendations generate 38% more click through ◦ Amazon: 35% sales from recommendations • Metrics which Affected by Recommendation ◦ Activity Metrics: Increase in user retention ◦ Financial Metrics: Increase in sales ◦ Product Activity: Increase on number of unique items bought Source: compiled from Xavier Amatriain MLSS slides (2014) Background
9. • The “Recommender problem” ◦ Estimate a utility function that automatically predicts how a user will like an item. ◦ Based on: ▪ Past behavior ▪ Relations to other users ▪ Item similarity ▪ Context Source: compiled from Xavier Amatriain MLSS slides (2014) Background
10. The “Recommender problem” • Let C be set of all users and let S be set of all possible recommendable items • Let u be a utility function measuring the usefulness of item s to user c, i.e., u : C X S→R, where R is a totally ordered set • For each user c є C, we want to choose items s є S that maximize u. Utility is usually represented by rating but can be any function Source: compiled from Xavier Amatriain MLSS slides (2014) Background
11. Possible Solutions Source: compiled from Xavier Amatriain MLSS slides (2014) Background
12. Problems in Recommendation System: Long Tail Sales Background
13. Problems in Recommendation System • Some users never use some items Background
14. SVD-like for Recommender System: Matrix Factorization to model RecSys
15. Non-Personalized Recommendation
16. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation a. For example, Content based filtering, recommend similar items. Product name embedding with Word2Vec SVD-like Recommendation
17. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation SVD-like Recommendation
18. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation a. Recommend items based on similarity: SVD-like Recommendation
19. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation a. Pros: i. Can be used if you don’t have any transaction history in the beginning ii. Sometimes can beat popular items benchmark. b. Cons: i. Low diversity metrics 1. Buy (mie-ayam), next recommendation (mie-ayam) SVD-like Recommendation
20. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation SVD-like Recommendation
21. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Non-Personalized Recommendation SVD-like Recommendation arg max similarity(item1, item2)
22. Personalized Recommendation
23. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 2. Personalized Recommendation SVD-like Recommendation arg max Probability(item, user)
24. Before dwelling to SVD-like RecSys CF, let’s focus on other approach: 1. Personalized Recommendation a. Based on items, users, and items-users interaction SVD-like Recommendation Alex Smola slides form Berkley ML class (2012)
25. How to model matrix completion: • Matrix Factorization SVD-like Recommendation Alex Smola slides form Berkley ML class (2012)
26. Matrix Factorization • Latent Variable SVD-like Recommendation Alex Smola slides form Berkley ML class (2012)
27. Matrix Factorization • Unfortunately we can’t user Matrix Factorization for sparse data SVD-like Recommendation Alex Smola slides form Berkley ML class (2012)
28. Funk SVD Alex Smola slides form Berkley ML class (2012)
29. Funk SVD Source: compiled from Alex Lin slides, CF with MF (2011)
30. SVD-like Recommender System Source: compiled from Alex Lin slides, CF with MF (2011)
31. SVD-like Recommender System Source: compiled from Alex Lin slides, CF with MF (2011)
32. SVD-like Recommender System Source: compiled from Alex Lin slides, CF with MF (2011)
33. SVD-like Recommender System Source: compiled from Alex Lin slides, CF with MF (2011)
34. Modeling Implicit and Explicit Feedback
35. Two types of Feedback Explicit and Implicit Feedback Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
36. In rating systems, such as MovieLens and Allmusic, the value of an entry denotes the rating of the item given by the user. Ratings can explicitly reflect the preference of an individual. Explicit and Implicit Feedback Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
37. For instance, in a music podcast service, the value of an entry may denote the subscription. Explicit and Implicit Feedback Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
38. Or play count Explicit and Implicit Feedback Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
39. Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
40. SVD-like model is so easy to fit any feedback type 1. Explicit a. Rating (Netflix Competition) Change the loss function to RMSE, regression task. Explicit and Implicit Feedback Source: compiled from Li Yen Kuo slides, Implicit RecSys (2019)
41. SVD-like model is so easy to fit any feedback type 1. Explicit b. Like/Dislike (Netflix Now) Change the loss function to log-loss, sigmoid activation. Explicit and Implicit Feedback
42. SVD-like model is so easy to fit any feedback type 2. Implicit a. Read/Not Read (Quora) Change loss function to BPR Loss Explicit and Implicit Feedback
43. SVD-like model is so easy to fit any feedback type 2. Implicit b. Frequency of Buy (Retail) Change loss function to Regression loss with Alpha, Koren (2009) Explicit and Implicit Feedback
44. Recommender System workflow: Estimate, Ranking and Randomization
45. Recommender System Workflow Source: compiled from Xavier Amatriain MLSS slides (2014)
46. Modeling 1. Build Recommender System based on Watch/Not a. Implicit model, maximize click through rate 2. Build Recommender System based on Rating a. Explicit model, maximize preference 3. Build Recommender System based on Churn a. Implicit model, maximize retention Recommender System Workflow Source: compiled from Xavier Amatriain MLSS slides (2014)
47. Recommender System Workflow Ranking • Most recommendations are presented in a sorted list • Recommendation can be understood as a ranking problem • Popularity is the obvious baseline • Ratings prediction is a clear secondary data input that allows for personalization • Many other features can be added
48. Recommender System Workflow Ranking Source: compiled from Xavier Amatriain MLSS slides (2014)
49. Randomization Randomization • We will focus on Diversity Metrics vs Accuracy Metrics. • More diverse recommendation will increase Netflix CTR • More accurate recommendation will increase Netflix CTR • Diversity and Accuracy are negatively correlated.
50. Randomization
51. Recommender System Cases
52. Questions Pair: Similarity Task
53. Questions Pair: Similarity Task Question 1 Question 2 {0, 1}
54. Questions Pair: Similarity Task • Is Item A similar to Item B?
55. Questions Pair: Similarity Task Input 1 Input 2 Neural Networks {0, 1}
56. Temporal Recommendation Task
57. Temporal Recommendation Task Item 1 Item 4 Item 2 Item 1 Item 2 Item 3 Item 2 Item 1 Item 5 RNN
58. Recommendation Task
59. Recommendation Task Click data Hotel 33 Hotel 77 . . . Hotel 44
60. Searching Task 60 Pseudo code BPR For K times: 1. A = Sample random label positive 2. B = Sample random label negative 3. Dist = Model predict distance(A,B) 4. Update, make the distance farther.
61. Searching Task Loss Function: 61
62. Searching Task • Pointwise ◦ for each items, train a classifier / regressor on it to predict how relevant it is • Pairwise ◦ given a pair of documents, compare which one has the highest rank. • Listwise ◦ sort the entire list of documents ▪ Direct optimization of IR measures such as NDCG
63. Searching Task: 63
64. Cowok 1 Cowok 3 Cowok 2 Cewek A Cewek C Cewek B Cowok 4 Cewek D Matching Task
65. Matching Task Cowok 1 Cewek C Cewek D Cewek A Cewek B Cewek A Cowok 4 Cowok 2 Cowok 1 Cowok 3
66. Matching Task 66 Cewek 1 Cowok 2 {0, 1}
67. Matching Task Loss Function: 67
68. Matching Task Matching Algorithm: Gale-Shapley 68
69. cowok cowok cowok cewek cewek cewek cowok cewek Matching Task
70. Matching Task 70
71. You can learn all of these in our Advance ML CLass Contact: business@pacmannai.com
72. We believe everyone can build cool Recommendation Systems Contact: business@pacmannai.com
73. Upcoming Advance ML CLass Contact: business@pacmannai.com
75. Check our website: www.pacmann.ai Contact: business@pacmannai.com
76. Check our website: www.pacmann.ai Contact: business@pacmannai.com
77. Check our website: www.pacmann.ai Contact: business@pacmannai.com
78. Past Classes Contact: business@pacmannai.com
79. Pacmann AI Classes Quality State of the Art of Machine Learning Research Practical Skills Theoretical Understanding > 50 institutions 400++ alumni 6 Classes in the past business@pacmannai.com https://pacmann.ai
80. Previous Classes business@pacmannai.com https://pacmann.ai 61 participants 8 weeks 48 institutions
81. Previous Classes business@pacmannai.com https://pacmann.ai 59 participants 8 weeks 48 institutions
82. Previous Classes business@pacmannai.com https://pacmann.ai 44 participants 2 weeks 33 institutions
83. Previous Classes Facts & Figure business@pacmannai.com https://pacmann.ai Field of Work
84. Previous Participants business@pacmannai.com https://pacmann.ai
85. Contact Email: business@pacmannai.com Whatsapp Business: +62 812-8122-1707