Mohammad Khabbaz
Laks V. S. Lakshmanan
ABSTRACT
Recommender systems help users find their items of interest from large data collections with little effort. Collaborative filtering (CF) is one of the most popular approaches for making recommendations. While significant work has been done on improving accuracy of CF methods, some of the most popular CF approaches are limited in terms of scalability and efficiency. The size of data in modern recommender systems is growing rapidly in terms of both new users and items and new ratings. Item-based recommendation is one of the CF approaches used widely in practice. It computes and uses an item-item similarity matrix in order to predict unknown ratings. Previous works on item-based CF method confirm its usefulness in providing high quality top-k results. In this paper, we design a scalable algorithm for top-k recommendations using this method. We achieve this by probabilistic modeling of the similarity matrix. A unique challenge here is that the ratings that are aggregated to produce the aggregate predicted score for a user should be obtained from different lists for different candidate items and the aggregate function is non-monotone. We propose a layered architecture for CF systems that facilitates computation of the most relevant items for a given user. We design efficient top-k algorithms and data structures in order to achieve high scalability. Our algorithm is based on abstracting the key computation of a CF algorithm in terms of two operations -- probe and explore. The algorithm uses a cost-based optimization whereby we express the overall cost as a function of a similarity threshold and determine its optimal value for minimizing the cost. We empirically evaluate our theoretical results on a large real world dataset. Our experiments show our exact top-k algorithm achieves better scalability compared to solid baseline algorithms.
- G. Adomavicius et al. Towards the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions. IEEE TKDE 17(6), 2005, pp. 734--749. Google Scholar
Digital Library
- J. S. Breese et al. Empirical Analysis of Predictive Algorithms for Collaborative Filtering. UAI 1998, pp. 43--52. Google ScholarDigital Library
- K. Chang et al. Minimal probing: supporting expensive predicates for top-k queries. SIGMOD 2002, pp. 346--357. Google ScholarDigital Library
- A. Das et al. Google news personalization: Scalable online collaborative filtering. WWW 2007, pp. 271--280. Google ScholarDigital Library
- M. Deshpande et al. Item-Based Top-N Recommendation Algorithms. ACM TOIS. 22(1), 2004, pp. 143--177. Google ScholarDigital Library
- R. Fagin et al. Optimal Aggregation Algorithms for Middleware. Journal of Computer System Sciences. Vol. 66, 2003, pp. 614--656. Google ScholarDigital Library
- J. A. Konstan, Introduction to recommender systems. SIGMOD 2008, pp. 1373--1374. Google ScholarDigital Library
- Y. Koren. Tutorial on Recent Progress in Collaborative Filtering, RecSys 2008, pp. 333--334. Google ScholarDigital Library
- D. Knuth. The Art of Computer Programming. Vol. 3, Fourth Edition, Addison-Wesley 2005. Google ScholarDigital Library
- Y. Luo et al. SPARK: Top-k Keyword Query in Relational Databases. SIGMOD 2007, pp. 115--126. Google ScholarDigital Library
- P. Resnick et al. GroupLens: An Open Architecture for Collaborative Filtering of Netnews. CSCW 1994, pp. 175--186. Google ScholarDigital Library
- B. Sarvar et al. Item-Based Collaborative Filtering Recommendation Algorithms. WWW 2001, pp. 285--295. Google ScholarDigital Library
- P. Symeonidis et al. Collaborative Recommender Systems: Combining Effectiveness and Efficiency. Expert Syst. Appl. 34(4), 2008, pp. 2995--3013. Google ScholarDigital Library
- M. Khabbaz and L. V. S. Lakshmanan. TopRecs: Top-k Algorithms for Item-based Collaborative Filtering. Tech. Report. Dept. of Computer Science, UBC September 2010. http://www.cs.ubc.ca/~laks/topK-techReport.pdf.Google Scholar
- M. Theobald et al. Top-k Query Evaluation with Probabilistic Guarantees. VLDB 2004, pp. 648--659. Google ScholarDigital Library
- D. Xin et al. Progressive and Selective Merge: Computing Top-K with Ad-hoc Ranking Functions. SIGMOD 2007, pp 103--114. Google ScholarDigital Library
- K. Yu et al. Instance Selection Techniques for Memory-Based Collaborative Filtering. SDM 2002.Google Scholar
- K. Yu et al. Probabilistic Memory-Based Collaborative Filtering. IEEE TKDE 16(1), 2004, pp. 56--69. Google ScholarDigital Library
- T. Wu et al. ARCube: Supporting Ranking Aggregate Quaries in Partially Materialized Data Cubes. SIGMOD 2008, pp. 79--92. Google ScholarDigital Library
- C. M. Bishop, Pattern Recognition and Machine Learning, Springer, 2006. Google ScholarDigital Library
- S. Amer-Yahia et al. Group Recommendation: Semantics and Efficiency. PVLDB 2(1), 2009, pp. 754--765. Google ScholarDigital Library
- G. Koutrika et al. FlexRecs: expressing and combining flexible recommendations. SIGMOD 2009, pp. 745--758. Google ScholarDigital Library
Index Terms
- TopRecs: Top-k algorithms for item-based collaborative filtering
Recommendations
A novel user-based collaborative filtering method by inferring tag ratings
User-based collaborative filtering is one of the most widely-used recommendation methods. It recommends items to a user based on her similar users' preferences. The essential part of user-based collaborative filtering is to infer users' similarities. A ...
User preference representation based on psychometric models
ADC '11: Proceedings of the Twenty-Second Australasian Database Conference - Volume 115Neighbourhood-based collaborative filtering is one of the most popular recommendation techniques, and has been applied successfully in various fields. User ratings are often used by neighbourhood-based collaborative filtering to compute the similarity ...
Using inferred tag ratings to improve user-based collaborative filtering
SAC '12: Proceedings of the 27th Annual ACM Symposium on Applied ComputingUser-based collaborative filtering is one of the most widely-used recommender methods. It recommends items to a user according to her similar users' opinions. The key point of user-based collaborative filtering is to compute users' similarities. In ...
Comments