PPT-Collaborative Filtering

Agenda Collaborative Filtering CF Pure CF approaches Userbased nearestneighbor The Pearson Correlation similarity measure Memorybased and modelbased approaches Itembased nearestneighbor. es Neal Lathia Dept of Computer Science University College of London Gower Street London WC1E 6BT UK nlathiacsuclacuk Josep M Pujol Telefonica Research Via Augusta 177 Barcelona 08021 Spain jmpstides Haewoon Kwak KAIST Computer Science Dept Kuseongdo 1600 Amphitheatre Pkwy Mountain View CA 94043 abhinandangooglecom Mayur Datar Google Inc 1600 Amphitheatre Pkwy Mountain View CA 94043 mayurgooglecom Ashutosh Garg Google Inc 1600 Amphitheatre Pkwy Mountain View CA 94043 ashutoshgooglecom Shyam Raja 25 129 2 77 519 116 3 78 509 122 4 78 497 129 5 77 497 150 6 76 492 173 7 77 471 102 8 76 467 149 9 77 464 140 10 75 460 129 11 75 453 120 12 78 451 135 13 77 451 120 14 77 445 153 15 75 437 126 16 74 436 147 17 76 421 147 18 77 421 120 19 78 419 145 PCA facilitates dimensionality reduction for of64258ine clus tering of users and rapid computation of recommendations For a database of users standard nearestneighbor tech niques require processing time to compute recom mendations whereas Eigentaste Collaborative 64257ltering the most success ful recommendation approach makes recommendations based on past transactions and feedback from consumers sharing similar interests A major problem limiting the usefulness of collaborative 64257ltering is t 25 129 2 77 519 116 3 78 509 122 4 78 497 129 5 77 497 150 6 76 492 173 7 77 471 102 8 76 467 149 9 77 464 140 10 75 460 129 11 75 453 120 12 78 451 135 13 77 451 120 14 77 445 153 15 75 437 126 16 74 436 147 17 76 421 147 18 77 421 120 19 78 419 145 Information Retrieval in Practice. All slides ©Addison Wesley, 2008. Social Search. Social search . Communities. of users . actively participating. in the search process. Goes beyond classical search tasks. Deep Packet Inspection. Artyom. . Churilin. Tallinn University of Technology 2011. Web filtering & DPI. Web filtering (content control) . is a way control . what content is permitted to a . user. . Kinan Halloum . 1. Presented paper. 2. Deep content-based music recommendation . by van den Oord et al. NIPS 2013. Outline. Music Recommendation. Collaborative filtering. Weighted Matrix Factorization. All slides ©Addison Wesley, 2008. Social Search. Social search . Communities. of users . actively participating. in the search process. Goes beyond classical search tasks. Key differences. Users interact with the system. Fouhey. .. Let’s Take An Image. Let’s Fix Things. Slide Credit: D. Lowe. We have noise in our image. Let’s replace each pixel with a . weighted. average of its neighborhood. Weights are . filter kernel. Henning Lange, Mario . Bergés. , Zico Kolter. Variational Filtering. Statistical Inference. (Expectation Maximization, Variational Inference). Deep Learning. Dynamical Systems. Variational Filtering. Outline. Recap. SVD . vs. PCA. Collaborative filtering. aka Social recommendation. k-NN CF methods. classification. CF via MF. MF . vs. SGD . vs. ….. Dimensionality Reduction. and Principle Components Analysis: Recap. Introduction to Recommender Systems. Recommender systems: The task. Customer W. 2. Slides adapted from Jure Leskovec. Plays an Ella Fitzgerald song. What should we recommend next?. Thomas . Quella. Wikimedia Commons.