Webtech recommender systems_presentation
•
0 likes•157 views
Recommender systems aim to provide personalized recommendations of items like movies, music, or products based on a user's preferences. Collaborative filtering approaches make recommendations based on the ratings of other similar users, while content-based filtering relies on item attributes. Hybrid systems combine these methods. The Netflix Prize competition spurred research on collaborative filtering algorithms like memory-based and model-based approaches. However, data sparsity and new user problems remain challenges.
Webtech recommender systems_presentation
- 1. Recommender Systems Simona Dakova The slides are licensed under a 1 Web Technologies – Prof. Dr. Ulrik Schroeder – WS 2010/11 Creative Commons Attribution 3.0 License
- 2. Overview Motivation Netflix Prize Competition Collaborative filtering approaches Content-based techniques Hybrid recommenders Summary 2 Web Technologies
- 3. We live in information overload! “We are leaving the age of Information and entering the Age of Recommendation” - The Long Tail (Chris Anderson) 3 Web Technologies
- 4. They try to attract you! Netflix: 2/3 of the movies rented were recommended Google News: 38% more click-throughs Amazon: 35% sales from recommendations 4 Web Technologies
- 5. Why recommenders? Enhance e-commerce and boost sales Browsers into buyers Recommender vs. Search: Discover the items you are looking, match your preferences Limited list of results Personalize your website content to the profile of an individual user Discover interesting items Automated personalization Increase usage and satisfaction 5 Web Technologies
- 6. Netflix Prize Competition $1.000.000 - if you “only” improve existing system by 10%! Contest started in 2006 Annual progress prize $ 50.000 Gained great popularity in academic circles The Winner BellKor´s Pragmatic Chaos 10.5% improvement in July 2009 6 Web Technologies
- 7. Recommender System = ? Definition: Algorithms/Systems for information filtering attempting to recommend certain items the user might like Items: Advertising messages, Investment choices, Restaurants, Cafes, Music tracks, Movies, TV programs, Books, Cloths, Supermarket goods, Tags, News articles, Online mates, Research papers 7 Web Technologies
- 8. User Profiling Understand people´s needs and interests Explicit Data Collection Ask for rating of items Rank a set of items Ask for detailed information/feedback CON: not well received by users, not ubiquitous Implicit Data Collection Purchasing history Items viewed Navigational patterns Obtain list of watched/listened items Analyze social data CON: Privacy concerns 8 Web Technologies
- 9. Technology overview RECOMMENDERS Collaborative filtering Content-based Hybrid (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms 9 Web Technologies
- 10. Collaborative filtering (CF) RECOMMENDERS Collaborative Content-based Hybrid filtering (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms • prediction based on past ratings • learn a model from user’s ratings • compute similarities between • use the model to predict the users/items probabilistic rating of the active user on given item • make prediction according to the calculated weight (similarity) 10 Web Technologies
- 11. Memory-based CF Algorithms RECOMMENDERS Collaborative Content-based Hybrid filtering (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms 11 Web Technologies
- 12. Memory-based CF Algorithms Entire or sample of the user-item matrix Steps: 1. For the active user/item identify his neighbors Similarity computation Pearson correlation Vector cosine-based similarity 2. Neighborhood-based prediction/ Top-N Recommendation 12 Web Technologies
- 13. User-based vs. Item-based i1 i2 i3 i4 i5 i1 i2 i3 i4 i5 u1 5 8 7 8 u1 5 8 7 8 u2 10 1 u2 10 1 u3 2 10 9 9 u3 2 10 9 9 u4 2 9 9 10 u4 2 9 9 10 u5 1 5 1 u5 1 5 1 ua 2 9 10 ua 2 9 10 User-based = You may like it Item-based = You may like it because your “friends” liked it because you like similar items 13 Web Technologies
- 14. Model-based CF Algorithms RECOMMENDERS Collaborative Content-based Hybrid filtering (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms 14 Web Technologies
- 15. Model-Based CF Algorithms R r5 E r9 C r11 O r3 r3 M r4 M r7 Train r7 r4 E r8 r8 N D r2 r1 A T r6 I MODEL O N all ratings (only set of ratings) Train your system to recognize complex patterns in user- item data (ratings) Make the recommendation based on the trained model Relies on machine learning and data mining algorithms 15 Web Technologies
- 16. Limitations and problems of CF Depend on human ratings Data sparsity Cold start , New user and New item problem Scalability Synonymy Shilling attacks Gray/Black sheep 16 Web Technologies
- 17. Content-based recommenders RECOMMENDERS Collaborative filtering Content-based Hybrid (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms 17 Web Technologies
- 18. Content-based recommendation (CB) For items containing textual information (keywords) Information Retrieval Compares similarity of the features of given items Example: Movie recommendation application Analyze common features among the movies Recommend only the movies that have a high degree of similarity to whatever the user’s preferences are Small Large Similarity SImilarity 18 Web Technologies
- 19. Limitations and problems of CB Limited content analysis Explicitly associated features Multimedia data – relies on tagging Same set of features – indistinguishable Overspecialization Difficult to recognize synonyms, concepts, or new emergi ng words New user Problem 19 Web Technologies
- 20. Hybrid recommenders RECOMMENDERS Collaborative filtering Content-based Hybrid (CF) Filtering (CB) recommenders Memory-based Model-based CF Algorithms CF Algorithms 20 Web Technologies
- 21. Hybrid recommenders Use combination of CF and CB Implementing methods separately and combining their predictions Incorporating CB characteristics into a CF approach or vice versa Constructing a general unifying model that incorporates both Example: content-boosted collaborative filtering i1 i2 i3 i4 i1 i2 i3 i4 Content Collaborative u1 5 8 x 7 u1 5 8 7 7 predictor filtering u2 10 x 1 x u2 10 4 1 8 RECOMMENDATION u3 2 x 10 9 u3 2 5 10 9 u4 x 2 9 9 u4 6 2 9 9 ua 2 x 9 10 ua 2 3 9 10 21 Web Technologies
- 22. Pros/Cons of Hybrid Recommenders Advantages Address limitations of pure CF or CB systems Provide more accurate recommendations Performance improvement Overcome sparsity Disadvatages Comlexity Expensive to build 22 Web Technologies
- 23. The winning solution on Netflix Contest A blend of several complex algorithms into a hybrid recommender system Main improvement: Incorporate temporal effects that cause movie and user biases as well as the changing user preferences 23 Web Technologies
- 24. Summary Techniques Advantages Limitations Memory-based algorithms: • easy implementation •data sparsity Neighborhood-based CF • no content considered •cold start problem Collaborative Top-N recommendation •limited scalability Model-based algorithms: • deal better with sparsity, • expensive modeling machine learning / data scalability • trade-off between mining algorithms • intuitive rationale performance and scalability Information retrieval • no data about other users • limited content Content-based • recommendation for analysis new/unpopular items • overspecialization • predictions for users with •new user problem unique tastes combination of • overcome limitations of pure • complexity collaborative and content- collaborative and content-based • expensive to build Hybrids based approaches recommendations • more accurate recommendations • performance improvement 24 Web Technologies
- 25. Literature Adomavicius, G., Tuzhilin, A. 2005. Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions. Su, X., Khoshgoftaar, T. 2009 A Survey on Collaborative Filtering Techniques. Sarwar, B., Karypis, G., Konstan, J., Riedl, J. 2001 Item-based collaborative Filtering Recommendation Algorithms. Das, A., Datar, M., Garg, A. 2007 Google News Personalization: Scalable Online Collaborative Fitlering. Linden, G., Smith, B., York, J. 2003 Amazon.com Recommendations Item-to- Item Collaborative Filtering. Guy, I., Zwerdling, N., Ronen, I., Carmel, D., Erel, U. 2010 Social Media Recommendation based on People and Tags. Schafer, J., Konstan, J., Riedl, J. 1999 Recommender Systems in E-Commerce. http://www.irelaxa.com/Geecat/2010/09/16/recommendation-system- collaborative-filtering/ Piotte, M., Chabbert, M. 2009 Extending the toolbox. 25 Web Technologies