A unified approach to building hybrid recommender systems

ABSTRACT

Content-based recommendation systems can provide recommendations for "cold-start" items for which little or no training data is available, but typically have lower accuracy than collaborative filtering systems. Conversely, collaborative filtering techniques often provide accurate recommendations, but fail on cold start items. Hybrid schemes attempt to combine these different kinds of information to yield better recommendations across the board.

We describe unified Boltzmann machines, which are probabilistic models that combine collaborative and content information in a coherent manner. They encode collaborative and content information as features, and then learn weights that reflect how well each feature predicts user actions. In doing so, information of different types is automatically weighted, without the need for careful engineering of features or for post-hoc hybridization of distinct recommender systems.

We present empirical results in the movie and shopping domains showing that unified Boltzmann machines can be used to combine content and collaborative information to yield results that are competitive with collaborative techniques in recommending items that have been seen before, and also effective at recommending cold-start items.

REFERENCES

• 1: D. Agarwal and S. Merugu. Predictive discrete latent factor models for large scale dyadic data. In SIGKDD, 2007.
• 2: C. Basu, H. Hirsh, and W. Cohen. Recommendation as classification: Using social and content-based information in recommendation. In AAAI, 1998.
• 3: J. Besag. Statistical analysis of non-lattice data. The Stat., 24(3):179--195, Sept. 1975.
• 4: D. Billsus and M. J. Pazzani. User modeling for adaptive news access. User Model. User-Adapt. Interact., 10(2-3):147--180, 2000.
• 5: J. S. Breese, D. Heckerman, and C. Kadie. Empirical analysis of predctive algorithms for collaborative filtering. In UAI, 1998.
• 6: R. D. Burke. Hybrid recommender systems: Survey and experiments. User Model. User-Adapt. Interact., 12(4):331--370, 2002.
• 7: S. F. Chen and R. Rosenfeld. A Gaussian prior for smoothing maximum entropy models. Technical Report CMU-CS-99-108, Carnegie Mellon University, 1999.
• 8: M. Claypool, A. Gokhale, T. Miranda, P. Murnikov, D. Netes, and M. Sartin. Combining content-based and collaborative filters in an online newspaper. In SIGIR Wkshp. Rec. Sys., 1999.
• 9: N. Good, J. B. Schafer, J. A. Konstan, A. Borchers, B. Sarwar, J. Herlocker, and J. Riedl. Combining collaborative filtering with personal agents for better recommendations. In AAAI, 1999.
• 10: A. Gunawardana and C. Meek. Tied Boltzmann machines for cold start recommendations. In ACM RecSys, 2008.
• 11: D. Heckerman, D. M. Chickering, C. Meek, R. Rounthwaite, and C. Kadie. Dependency networks for inference, collaborative filtering, and data visualization. JMLR, 1:49--75, Oct. 2000.
• 12: T. Hofmann and J. Puzicha. Latent class models for collaborative filtering. In IJCAI, 1999.
• 13: C.-N. Hsu, H.-H. Chung, and H.-S. Huang. Mining skewed and sparse transaction data for personalized shopping recommendation. Mach. Learn., 57(1-2):35--59, 2004.
• 14: A. Hyvarinen. Consistency of pseudolikelihood estimation of fully visible Boltzmann machines. Neur. Comp., 18:2283--2292, 2006.
• 15: M. I. Jordan, Z. Ghahramani, T. S. Jaakkola, and L. K. Saul. An introduction to varational methods for graphical models. In M. I. Jordan, editor, Learning in Graphical Models, pages 105--162. MIT Press, 1999.
• 16: M. Kurucz, A. A. Benczur, and K. Csalogany. Methods for large scale SVD with missing values. In KDD Cup, 2007.
• 17: G. Linden, B. Smith, and J. York. Amazon.com recommendations: Item-to-item collaborative filtering. IEEE Internet Comp. Mag., 7(1), 2003.
• 18: R. F. Lyon and L. S. Yaeger. On-line hand-printing recognition with neural networks. In MicroNeuro, 1996.
• 19: A. McCallum and K. Nigam. A comparison of event models for naive bayes text classification. In AAAI Wkshp. Lrn. Txt. Cat., 1998.
• 20: M. Montaner, B. Lopez, and J. L. de la Rosa. A taxonomy of recommender agents on the internet. AI Rev., 19(4), 2003.
• 21: R. J. Mooney and L. Roy. Content-based book recommending using learning for text categorization. In ACM DL, pages 195--204, 2000.
• 22: S.-T. Park, D. Pennock, O. Madani, N. Good, and D. DeCoste. Na¨1ve filterbots for robust cold-start recommendations. In SIGKDD, pages 699--705, 2006.
• 23: M. J. Pazzani. A framework for collaborative, content-based and demographic filtering. AI Rev., 13:393--408, 1999.
• 24: P. Resnick, N. Iacovou, M. Suchak, P. Bergstrom, and J. Riedl. GroupLens: an open architecture for collaborative filtering of netnews. In ACM CSCW, pages 175--186, 1994.
• 25: M. Riedmiller and H. Braun. A direct adaptive method for faster backpropagation learning: The RPROP algorithm. In IEEE Int. Conf. Neur. Net., volume 1, pages 586--591, 1993.
• 26: R. Salakhutdinov. Learning and evaluating boltzmann machines. Technical Report UTML TR 2008-002, Dept. of Computer Science, Univ. of Toronto, 2008.
• 27: R. Salakhutdinov, A. Mnih, and G. Hinton. Restricted Boltzmann machines for collaborative filtering. In ICML, 2007.
• 28: B. Sarwar, G. K. nad Joseph Konstan, and J. Reidl. Item-based collaborative filtering recommendation algorithms. In WWW, 2001.
• 29: A. I. Schein, A. Popescul, L. H. Ungar, and D. M. Pennock. Generative models for cold-start recommendations. In SIGIR Wkshp. Recom. Sys., 2001.
• 30: A. I. Schein, A. Popescul, L. H. Ungar, and D. M. Pennock. Methods and metrics for cold-start recommendations. In SIGIR. ACM, Aug. 2002.
• 31: A. Zimdars, D. M. Chickering, and C. Meek. Using temporal data for making recommendations. In UAI, 2001.

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.3 Information Search and Retrieval
          Subjects: Information filtering

Additional Classification:
  G. Mathematics of Computing
  G.3 PROBABILITY AND STATISTICS
      Subjects: Correlation and regression analysis

  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.6 Learning
          Subjects: Parameter learning

General Terms:
Algorithms, Performance

Keywords:
boltzmann machines, cold start, collaborative filtering, content-based filtering, recommender systems