Collaborative Filtering With Decoupled Models for Preferences and Ratings Rong Jin 1, Luo Si 1, ChengXiang Zhai 2 and Jamie Callan 1 Language Technology.

Slides:



Advertisements
Similar presentations
Item Based Collaborative Filtering Recommendation Algorithms
Advertisements

Jeff Howbert Introduction to Machine Learning Winter Collaborative Filtering Nearest Neighbor Approach.
1 RegionKNN: A Scalable Hybrid Collaborative Filtering Algorithm for Personalized Web Service Recommendation Xi Chen, Xudong Liu, Zicheng Huang, and Hailong.
COLLABORATIVE FILTERING Mustafa Cavdar Neslihan Bulut.
Active Learning and Collaborative Filtering
The Wisdom of the Few A Collaborative Filtering Approach Based on Expert Opinions from the Web Xavier Amatriain Telefonica Research Nuria Oliver Telefonica.
2. Introduction Multiple Multiplicative Factor Model For Collaborative Filtering Benjamin Marlin University of Toronto. Department of Computer Science.
Memory-Based Recommender Systems : A Comparative Study Aaron John Mani Srinivasan Ramani CSCI 572 PROJECT RECOMPARATOR.
Lecture 14: Collaborative Filtering Based on Breese, J., Heckerman, D., and Kadie, C. (1998). Empirical analysis of predictive algorithms for collaborative.
1 Collaborative Filtering Rong Jin Department of Computer Science and Engineering Michigan State University.
Modeling User Rating Profiles For Collaborative Filtering
Recommender systems Ram Akella February 23, 2011 Lecture 6b, i290 & 280I University of California at Berkeley Silicon Valley Center/SC.
Recommender systems Ram Akella November 26 th 2008.
Algorithms for Efficient Collaborative Filtering Vreixo Formoso Fidel Cacheda Víctor Carneiro University of A Coruña (Spain)
Federated Search of Text Search Engines in Uncooperative Environments Luo Si Language Technology Institute School of Computer Science Carnegie Mellon University.
1 Collaborative Filtering: Latent Variable Model LIU Tengfei Computer Science and Engineering Department April 13, 2011.
Collaborative Filtering & Content-Based Recommending
Learning Table Extraction from Examples Ashwin Tengli, Yiming Yang and Nian Li Ma School of Computer Science Carnegie Mellon University Coling 04.
Combining Content-based and Collaborative Filtering Department of Computer Science and Engineering, Slovak University of Technology
Chapter 12 (Section 12.4) : Recommender Systems Second edition of the book, coming soon.
Item-based Collaborative Filtering Recommendation Algorithms
A NON-IID FRAMEWORK FOR COLLABORATIVE FILTERING WITH RESTRICTED BOLTZMANN MACHINES Kostadin Georgiev, VMware Bulgaria Preslav Nakov, Qatar Computing Research.
Distributed Networks & Systems Lab. Introduction Collaborative filtering Characteristics and challenges Memory-based CF Model-based CF Hybrid CF Recent.
1 Information Filtering & Recommender Systems (Lecture for CS410 Text Info Systems) ChengXiang Zhai Department of Computer Science University of Illinois,
1 Formal Models for Expert Finding on DBLP Bibliography Data Presented by: Hongbo Deng Co-worked with: Irwin King and Michael R. Lyu Department of Computer.
A Hybrid Recommender System: User Profiling from Keywords and Ratings Ana Stanescu, Swapnil Nagar, Doina Caragea 2013 IEEE/WIC/ACM International Conferences.
Classical Music for Rock Fans?: Novel Recommendations for Expanding User Interests Makoto Nakatsuji, Yasuhiro Fujiwara, Akimichi Tanaka, Toshio Uchiyama,
Chengjie Sun,Lei Lin, Yuan Chen, Bingquan Liu Harbin Institute of Technology School of Computer Science and Technology 1 19/11/ :09 PM.
Presented By :Ayesha Khan. Content Introduction Everyday Examples of Collaborative Filtering Traditional Collaborative Filtering Socially Collaborative.
Google News Personalization: Scalable Online Collaborative Filtering
1 Recommender Systems Collaborative Filtering & Content-Based Recommending.
Developing Trust Networks based on User Tagging Information for Recommendation Making Touhid Bhuiyan et al. WISE May 2012 SNU IDB Lab. Hyunwoo Kim.
EigenRank: A Ranking-Oriented Approach to Collaborative Filtering IDS Lab. Seminar Spring 2009 강 민 석강 민 석 May 21 st, 2009 Nathan.
Collaborative Filtering  Introduction  Search or Content based Method  User-Based Collaborative Filtering  Item-to-Item Collaborative Filtering  Using.
The Effect of Dimensionality Reduction in Recommendation Systems
Evaluation of Recommender Algorithms for an Internet Information Broker based on Simple Association Rules and on the Repeat-Buying Theory WEBKDD 2002 Edmonton,
A Content-Based Approach to Collaborative Filtering Brandon Douthit-Wood CS 470 – Final Presentation.
1 Collaborative Filtering & Content-Based Recommending CS 290N. T. Yang Slides based on R. Mooney at UT Austin.
EigenRank: A ranking oriented approach to collaborative filtering By Nathan N. Liu and Qiang Yang Presented by Zachary 1.
Carnegie Mellon Novelty and Redundancy Detection in Adaptive Filtering Yi Zhang, Jamie Callan, Thomas Minka Carnegie Mellon University {yiz, callan,
Cosine Similarity Item Based Predictions 77B Recommender Systems.
Collaborative Filtering Zaffar Ahmed
Effective Automatic Image Annotation Via A Coherent Language Model and Active Learning Rong Jin, Joyce Y. Chai Michigan State University Luo Si Carnegie.
Pearson Correlation Coefficient 77B Recommender Systems.
Collaborative Filtering via Euclidean Embedding M. Khoshneshin and W. Street Proc. of ACM RecSys, pp , 2010.
ICONIP 2010, Sydney, Australia 1 An Enhanced Semi-supervised Recommendation Model Based on Green’s Function Dingyan Wang and Irwin King Dept. of Computer.
Personalization Services in CADAL Zhang yin Zhuang Yuting Wu Jiangqin College of Computer Science, Zhejiang University November 19,2006.
Online Evolutionary Collaborative Filtering RECSYS 2010 Intelligent Database Systems Lab. School of Computer Science & Engineering Seoul National University.
User Modeling and Recommender Systems: recommendation algorithms
CYUT ISKM 2004/01/13 1 Fuzzy logic methods in recommender systems Author: Ronald R. Yager Source:Fuzzy set and systems, Vol. 134, 2003, pp Presented.
The Effect of Database Size Distribution on Resource Selection Algorithms Luo Si and Jamie Callan School of Computer Science Carnegie Mellon University.
Relevant Document Distribution Estimation Method for Resource Selection Luo Si and Jamie Callan School of Computer Science Carnegie Mellon University
Reputation-aware QoS Value Prediction of Web Services Weiwei Qiu, Zhejiang University Zibin Zheng, The Chinese University of HongKong Xinyu Wang, Zhejiang.
Collaborative Deep Learning for Recommender Systems
ItemBased Collaborative Filtering Recommendation Algorithms 1.
Slope One Predictors for Online Rating-Based Collaborative Filtering Daniel Lemire, Anna Maclachlan In SIAM Data Mining (SDM’05), Newport Beach, California,
Item-Based Collaborative Filtering Recommendation Algorithms
Designing a framework For Recommender system Based on Interactive Evolutionary Computation Date : Mar 20 Sat, 2011 Project Number :
Tagommenders: Connecting Users to Items through Tags Written by Shilad Sen, Jesse Vig, and John Riedl (2009) Presented by Ken Hu and Hassan Hattab.
Data Mining: Concepts and Techniques
Sampath Jayarathna Cal Poly Pomona
WSRec: A Collaborative Filtering Based Web Service Recommender System
Methods and Metrics for Cold-Start Recommendations
Asymmetric Correlation Regularized Matrix Factorization for Web Service Recommendation Qi Xie1, Shenglin Zhao2, Zibin Zheng3, Jieming Zhu2 and Michael.
Adopted from Bin UIC Recommender Systems Adopted from Bin UIC.
Collaborative Filtering Nearest Neighbor Approach
M.Sc. Project Doron Harlev Supervisor: Dr. Dana Ron
Movie Recommendation System
ITEM BASED COLLABORATIVE FILTERING RECOMMENDATION ALGORITHEMS
Response Aware Model-Based Collaborative Filtering
Presentation transcript:

Collaborative Filtering With Decoupled Models for Preferences and Ratings Rong Jin 1, Luo Si 1, ChengXiang Zhai 2 and Jamie Callan 1 Language Technology Inst School of Computer Science Carnegie Mellon University 1 {rong, lsi, Dept of Computer Science University of Illinois at Urbana-Champaign 2

2 © CIKM 2003 Abstract Task: New algorithm to address an important problem of collaborative filtering systems and to improve the performance Outline: Introduction to collaborative filtering Previous work Decoupled Model (DM) of decoupling user preferences and ratings Experiment results Related work Conclusion and future work

3 © CIKM 2003 What is Collaborative Filtering? Collaborative Filtering (CF): Making recommendation decisions for a specific user based on the judgments of users with similar tastes Content-Based Filtering: Recommend by analyzing the content information Collaborative Filtering: Make recommendation by judgments of similar users

4 © CIKM 2003 Why Collaborative Filtering? Advantages of Collaborative Filtering: The contents of items belong to the third-party (not accessible or available) The contents of items are difficult to index or analyze (multimedia information etc) Applications:

5 © CIKM 2003 Formal Framework for Collaborative Filtering Test User U t 2 3 What we have: Assume there are some ratings by training users Test user provides some amount of additional training data What we do: Predict test user’s rating based training information R u t (O j ) = Training Users: U n O 1 O 2 O 3 ……O j ………… O M U1U2U1U2 UNUN UiUi Objects: O m 3 2 4

6 © CIKM 2003 Previous Work: Memory-Based Approaches Memory-Based Approaches: No training procedure Calculate similarities of training users to test user and predict with weighted average of training users’ ratings Pearson Correlation Coefficient Similarity Average Ratings Vector Space Similarity Prediction:

7 © CIKM 2003 Previous Work: Model-Based Approaches Model-Based Approaches: Aspect Model (Hofmann et al., 1999) –Model individual ratings as convex combination of preference factors Z R O U P(o|Z) P(Z) P(u|Z) P(r|Z) Personality Diagnosis Model (Pennock et al., 1999) –Hybrid of memory and model-based approach

8 © CIKM 2003 Previous Work: Thoughts Thoughts: Previous algorithms address the problem that users with similar tastes may have different rating patterns implicitly (Normalize user rating) Explicitly decouple users preference values out of the rating values Decoupled Model (DM) Nice Rating: 5 Mean Rating: 2 Nice Rating: 3 Mean Rating: 1

9 © CIKM 2003 Decoupled Model (DM) Decoupled Model (DM): Task: Separate preference values out of surface rating values Preference Value PV= PV=0.667 (0 disfavor,1 favor) PV=

10 © CIKM 2003 Decoupled Model (DM) Simple method User rating frequency vector Smoothed version User Rating Pattern Similarity

11 © CIKM 2003 Decoupled Model (DM) Memory-Based approaches with preference values Predict preference value on an object of test user User Preference Pattern Similarity Convert preference value back to rating value

12 © CIKM 2003 Experimental Data MovieRatingEachMovie Number of Users Number of Movies Avg. # of rated items/User Scale of ratings1,2,3,4,51,2,3,4,5,6 Datasets: MovieRating and EachMovie Evaluation: MAE: average absolute deviation of the predicted ratings to the actual ratings on items.

13 © CIKM 2003 Experimental Methodology Vary Number of Training User Test behaviors of algorithms with different amount of training data –For MovieRating 100 and 200 training users –For EachMovie 200 and 400 training users Vary Amount of Given Information from the Test User Test behaviors of algorithms with different amount of given information from test user –For both testbeds Vary among given 5, 10, or 20 items

14 © CIKM 2003 Experimental Results New Model and Other Baseline Algorithms Movie Rating, 100 Training Users Movie Rating, 200 Training Users Each Movie, 400 Training Users Each Movie, 200 Training Users Given: Given: Given: Given: MAEMAE MAEMAE MAEMAE MAEMAE PCC VS PD AM New Model

15 © CIKM 2003 Experimental Results Compare Two Methods of Computing Preference Values Training Users Size Algorithms 5 Items Given 10 Items Given 20 Items Given 100 Simple Smoothed Simple Smoothed Results on Movie Rating Training Users Size Algorithms 5 Items Given 10 Items Given 20 Items Given 200 Simple Smoothed Simple Smoothed Results on Each Movie

16 © CIKM 2003 Decoupled Model (DM) Simple method User rating frequency vector Smoothed version User Rating Pattern Similarity

17 © CIKM 2003 Experimental Results Compare Two Methods of Computing Preference Values Training Users Size Algorithms 5 Items Given 10 Items Given 20 Items Given 100 Simple Smoothed Simple Smoothed Results on Movie Rating Training Users Size Algorithms 5 Items Given 10 Items Given 20 Items Given 200 Simple Smoothed Simple Smoothed Results on Each Movie

18 © CIKM 2003 Conclusion and Future Work Conclusions: Propose the decoupled model –Explicitly extract preference values from the surface rating values –Combine the decoupled model with memory-based approach and improve the performance Our Related Work: Combine decoupled model with model-based approach (ICML’03) A more formal and unified probabilistic graphical model (UAI’03) Future Work: Combine content-based filtering and collaborative filtering recommendation methods together.