Applying HOSVD to Alleviate the Sparsity Problem in Context-aware Recommender Systems

WANG Licai; MENG Xiangwu; ZHANG Yujie

Article Contents

Article Navigation > Chinese Journal of Electronics > 2013 > 22(4): 773-778

WANG Licai, MENG Xiangwu, ZHANG Yujie, “Applying HOSVD to Alleviate the Sparsity Problem in Context-aware Recommender Systems,” Chinese Journal of Electronics, vol. 22, no. 4, pp. 773-778, 2013,

Citation:

WANG Licai, MENG Xiangwu, ZHANG Yujie, “Applying HOSVD to Alleviate the Sparsity Problem in Context-aware Recommender Systems,” Chinese Journal of Electronics, vol. 22, no. 4, pp. 773-778, 2013,

WANG Licai, MENG Xiangwu, ZHANG Yujie, “Applying HOSVD to Alleviate the Sparsity Problem in Context-aware Recommender Systems,” Chinese Journal of Electronics, vol. 22, no. 4, pp. 773-778, 2013,

Citation:

PDF( 265 KB)

Applying HOSVD to Alleviate the Sparsity Problem in Context-aware Recommender Systems

WANG Licai^{1,2
,},
MENG Xiangwu^1,2,
ZHANG Yujie^1,2

1.
School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China;
2.
Beijing Key Laboratory of Intelligent Telecommunications Software and Multimedia, Beijing 100876, China

Funds: This work is supported by the National Natural Science Foundation of China (No.60872051), the Fundamental Research Funds for the Central Universities (No.2009RC0203), and the Mutual Project of Beijing Municipal Education Commission of China.

More Information

Corresponding author: WANG Licai, MENG Xiangwu, ZHANG Yujie
Received Date: 2012-06-01
Rev Recd Date: 2012-10-01
Publish Date: 2013-09-25

Abstract

Abstract

It is quite a great challenge for Contextaware recommender systems (CARS) to generate accurate recommendations with only a few available none-zero contextual user preferences. This paper presents a new approach to alleviate this high sparsity problem by applying the Higher order singular value decomposition (HOSVD) technique. Firstly, it constructs an N-order tensor to represent multidimensional contextual user preferences and decompose it into (N-2) 3-order tensors according to different types of context (such as time, location and activity). Secondly, it introduces HOSVD to automatically discover the latent associations among these different 3dimensional objects and predicts unknown unidimensional contextual user preferences. Finally, it calculates every contextual influence coefficient that each type of context factor influences user preferences and then constructs a new N-order tensor using weighted linearization method in order to provide recommendations. Experimental evaluation on a simulated personalized mobile services environment demonstrates the efficacy of our approach against the other baseline ones.
- Context-aware recommender systems,
- Sparsity,
- User preference,
- Contextual information,
- Higher order singular value decomposition (HOSVD)

FullText(HTML)

References(16)

References

G. Adomavicius, A. Tuzhilin, “Toward the next generation of recommender systems: A survey of the state-of-the-art and possiblee xtensions”, IEEE Transactions on Knowledge and DataE ngineering, Vol.17, No.2, pp.734-749, 2005.

J. Cao et al., “A novel collaborative filtering using kernel methodsf or recommender systems”, Chinese Journal of Electronics,Vol.21, No.4, pp.609-614, 2012.

L.C. Wang, X.W. Meng, Y.J. Zhang, “Context-aware recommenders ystems”, Chinese Journal of Software, Vol.23, No.1,p p.1-20, 2012.

G. Adomavicius, R. Sankaranarayanan, S. Sen, A. Tuzhilin, “Incorporatingc ontextual Information in recommender systems usinga multidimensional approach”, ACM Transactions on InformationS ystems, Vol.23, No.1, pp.103-145, 2005.

Z. Huang, H. Chen, D. Zeng, “Applying associative retrievalt echniques to alleviate the sparsity problem in collaborative filtering”,A CM Transactions on Information Systems, Vol.22,N o.1, pp.116-142, 2004.

H. Yildirim, M.S. Krishnamoorthy, “A random walk method fora lleviating the sparsity problem in collaborative filtering”, Proc. of ACM Recsys'08, Switzerland, pp.131-138, 2008.

D. Billsus, M. Pazzani, “Learning collaborative information filters”, Proc. of the 15th International Conference on MachineL earning, San Francisco, USA, pp.46-53, 1998.

V. Sindhwani et al., “A family of non-negative matrix factorizationsf or one-class collaborative filtering”, Proc. of ACM Recsys'0 9: Recommender based Industrial Applications Workshop,New York, USA, pp.1-8, 2009.

Y. Koren, “Factor in the neighbors: Scalable and accurate collaborativef iltering”, ACM Transactions on Knowledge Discovery from Data, Vol.4, No.1, pp.1-24, 2010.

T.Z. Zhang, R.M. Shen, H.T. Lu, “Using non-negative matrixf actorization to cluster learners and construct learning communities”, Chinese Journal of Electronics, Vol.20, No.2, pp.207-2 11, 2011.

De L. Lathauwer, De B. Moor, J. Vandewalle, “A multilinears ingular value decomposition”, SIAM Journal on Matrix Analysis and Applications, Vol.21, No.4, pp.1253-1278, 2000.

J. Sun et al., “CubeSVD: A novel approach to personalizedw eb search”, Proc. of the Fourteenth International World WideW eb Conference, Chiba, Japan, pp.382-390, 2005.

P. Symeonidis, A. Nanopoulos, Y. Manolopoulos, “A unifiedf ramework for providing recommendations in social tagging systemsb ased on ternary semantic analysis”, IEEE Transactionso n Knowledge and Data Engineering, Vol.22, No.2, pp.179-192,2 010.

V. Zheng et al., “Collaborative filtering meets mobile recommendation:A user-centered approach”, Proc. of the Twenty-F ourth AAAI Conference on Artificial Intelligence, Atlanta,G eorgia, USA, pp.236-241, 2010.

J. Su, H. Yeh, P. Yu, V. Tseng, “Music recommendation usingc ontent and context information mining”, IEEE IntelligentS ystems, Vol.25, No.1, pp.16-26, 2010.

L.C. Wang, X.W. Meng, Y.J. Zhang, “A cognitive psychologybaseda pproach to user preferences elicitation for mobile networks ervices”, Acta Electronica Sinica, Vol.39, No.11, pp.2547-2 553, 2011.

Relative Articles

Supplements(0)

Cited By

Proportional views