Local Low-Rank Matrix Approximation with Preference Selection of Anchor Points

Matrix factorization is widely used in personalized recommender systems, text mining, and computer vision. A general assumption to construct matrix approximation is that the original matrix is of global low rank, while Joonseok Lee et al. proposed that many real matrices may be not globally low rank, and thus a locally low-rank matrix approximation method has been proposed.[11] However, this kind of matrix approximation method still leaves some important issues unsolved, for example, the randomly selecting anchor nodes. In this paper, we study the problem of the selection of anchor nodes to enhance locally low-rank matrix approximation. We propose a new model for local low-rank matrix approximation which selects anchor-points using a heuristic method. Our experiments indicate that the proposed method outperforms many state-of-the-art recommendation methods. Moreover, the proposed method can significantly improve algorithm efficiency, and it is easy to parallelize. These traits make it potential for large scale real-world recommender systems.

[1]  John Riedl,et al.  GroupLens: an open architecture for collaborative filtering of netnews , 1994, CSCW '94.

[2]  Sung Ho Ha,et al.  Helping Online Customers Decide through Web Personalization , 2002, IEEE Intell. Syst..

[3]  Philip S. Yu,et al.  Semantic Path based Personalized Recommendation on Weighted Heterogeneous Information Networks , 2015, CIKM.

[4]  Michael R. Lyu,et al.  Learning to recommend with social trust ensemble , 2009, SIGIR.

[5]  Wei Pang,et al.  Hete-CF: Social-Based Collaborative Filtering Recommendation Using Heterogeneous Relations , 2014, 2014 IEEE International Conference on Data Mining.

[6]  Yizhou Sun,et al.  Personalized entity recommendation: a heterogeneous information network approach , 2014, WSDM.

[7]  References , 1971 .

[8]  Alejandro Bellogín,et al.  A comparative study of heterogeneous item recommendations in social systems , 2013, Inf. Sci..

[9]  James Bennett,et al.  The Netflix Prize , 2007 .

[10]  Matthew P. Wand,et al.  Kernel Smoothing , 1995 .

[11]  Harald Steck,et al.  Circle-based recommendation in online social networks , 2012, KDD.

[12]  Jiawei Han,et al.  On building entity recommender systems using user click log and freebase knowledge , 2014, WSDM.

[13]  Yehuda Koren,et al.  Factorization meets the neighborhood: a multifaceted collaborative filtering model , 2008, KDD.

[14]  Samy Bengio,et al.  Local collaborative ranking , 2014, WWW.

[15]  Yoram Singer,et al.  Local Low-Rank Matrix Approximation , 2013, ICML.

[16]  Alejandro Bellogín,et al.  Content-based recommendation in social tagging systems , 2010, RecSys '10.

[17]  Michael R. Lyu,et al.  SoRec: social recommendation using probabilistic matrix factorization , 2008, CIKM '08.

[18]  Douglas B. Terry,et al.  Using collaborative filtering to weave an information tapestry , 1992, CACM.

[19]  Ruslan Salakhutdinov,et al.  Probabilistic Matrix Factorization , 2007, NIPS.

[20]  Yizhou Sun,et al.  Recommendation in heterogeneous information networks with implicit user feedback , 2013, RecSys.

[21]  Chao Liu,et al.  Recommender systems with social regularization , 2011, WSDM '11.

[22]  Jianyong Wang,et al.  Incorporating heterogeneous information for personalized tag recommendation in social tagging systems , 2012, KDD.

[23]  Philip S. Yu,et al.  Integrating heterogeneous information via flexible regularization framework for recommendation , 2015, Knowledge and Information Systems.

[24]  Mark Rosenstein,et al.  Recommending and evaluating choices in a virtual community of use , 1995, CHI '95.

[25]  Pattie Maes,et al.  Social information filtering: algorithms for automating “word of mouth” , 1995, CHI '95.