Recommender Systems David M Pennock NEC Research Institute

Recommender Systems David M. Pennock NEC Research Institute contributions: John Riedl, Group. Lens University of Minnesota

CW: scale vs. service 4 Wal-Mart + massive inventory + massive customer base + cheap – impersonal 4 General store – specialized products – few customers – expensive + knowledgeable • about products • about YOU

The vision of automation: Mass personalization 4 Wal-Mart. com + massive inventory + massive customer base + cheap – impersonal + knowledgeable • about products • about YOU

Commerce: Matching buyers and sellers

Commerce: Matching buyers and sellers Traditional: – – browsing ads critics/editors friends Technological facilitators: – World Wide Web – targeted ads – search engines/ shop bots – recommender systems

Research groups 4 Group. Lens @ University of Minnesota – Riedl, Konstan et al. – Movie. Lens; Net. Perceptions; tutorial 4 Microsoft Research – Breese, Heckerman, Horvitz, et al. – Site. Server; Firefly 4 MIT – Maes et al. ; Firefly 4 NEC Research, U. Penn – Pennock, Lawrence, Ungar, Popescul

Types of recommender systems Content-based information filter uses AI techniques A romantic comedy starring Julia Roberts in stock at BB A movie like Fargo

Types of recommender systems Community-based collaborative filter intelligence from people Hybrid systems A movie that people like me enjoyed

Collaborative filtering: How it works ratings Thanks: John Riedl & Group. Lens Ratings Correlations

Collaborative filtering: How it works 2 1 3 1 ratings Fargo = 2 Thanks: John Riedl & Group. Lens Ratings Correlations neighbors

Examples and applications 4 News 4 Movies: http: //www. movielens. umn. edu 4 Books 4 Websites: Alexa. com 4 Music, toys, … 4 Netperceptions. com 4 CDNow. com, Levis. com, … 4 Commerce Edition of Microsoft Site. Server

Group. Lens: Usenet news ‘ 94 Thanks: John Riedl & Group. Lens

Movie. Lens Thanks: John Riedl & Group. Lens

Amazon. com

800. com accessories (for browsers)

Launch. com

Cdnow album advisor

Jester

Ecommerce success stories 4 Large international catalog retailer – 17% hit rate, 23% acceptance rate in call center 4 Medium European outbound call center – 17% hit rate, 6. 7% acceptance rate from an outbound telemarketing call – $350. 00 price of average item sold – Items were in an electronics over-stocked category and were soldout within 3 weeks 4 Medium American online toy store (e-mail campaign) – 19% click-thru rate vs. 10% industry average – 14. 3% conversion to sale vs. 2. 5% industry average Thanks: John Riedl & Group. Lens

Algorithms: Memory-based 2 1 3 1 ratings Ra(Fargo) = i wi Ri(Fargo) for each movie neighbors where is over neighborhood (k-NN, k-radius); similarity metric wi is correlation, or vector similarity, or mean squared difference, or prob of same “personality” [Pennock et al. ], or… Group. Lens [Resnick et al. 94]; Ringo [Shardanand Maes 95]; comparative study [Breese et al. 98]

Algorithms: Model-based action teenage, male ratings Build underlying model of user preferences; infer predictions from model Personality diagnosis [Pennock, Horvitz, Lawrence, & Giles 2000] Bayesian network [Breese et al. 98] variables are products; values are ratings; structure and probs learned Bayesian clustering Like-minded users grouped [Breese et al. 98] Users and products clustered [Ungar and Foster 1998]

Algorithms: Machine learning Black box machine learning or classification problem: Ripper [Basu et al. 98] Neural network Support vector machine [Billsus and Pazzani 98; Freund et al. 98; Nakamura and Abe 98]

State of the art 4 Weighted k-nearest neighbor! 4 Singular value decomposition [Group. Lens] 4 Probabilistic SVD - Aspect model [Hofmann and Puzicha 99] [Popescul, Ungar, Pennock, and Lawrence] 4 Some problems/hurdles – data sparsity (one solution: smoothing) – implicit ratings (one solution: “boosting”) Thanks: John Riedl & Group. Lens • purchase history [Ungar] [Claypool] [Sarwar & Karypis] • access history/time spent reading [Morita and Shen] [Pennock et al. 2000] [Popescul et al. ]

Filtering content 4 Research. Index [Pennock et al. 2000] [Popescul et al. ] 4 Personalized news [Claypool et al. 99] 4 Personalized search engines – Beyond keyword search 4 Adaptive web sites [Etzioni et al. ] 4 Justifying subscriptions Thanks: John Riedl & Group. Lens

GOOGLE

Extensions 4 Incorporating content, links, other data – Filter. Bots [Group. Lens] – Ripper [Basu et al. 98] – three-way aspect model [Popescul et al. ] 4 Group recommendations 4 Temporal aspects 4 “schizophrenic” users – moods / changing and “ephemeral” tastes – buying for others

Multiuser, from movielens

Conclusion 4 Mass personalization – expensive or impossible without automation – large retailers act and “feel” small Thanks: John Riedl & Group. Lens 4 Recommender systems – intelligence from leveraging community information, rather than just AI – can incorporate content, demographic information, etc. – can scale to millions of customers, millions of products, thousands of clicks per second – ideally adds value for both retailers & consumers