Xiaoxi Shi
ABSTRACT
Collaborative Filtering algorithm is widely used in plentiful personal recommendation system. However, it has low accuracy prediction in sparse data set. Current mainstream collaborative filtering algorithm filter neighbor of target user by calculating similarity between users with co-rated ratings. Nonnegative Matrix factorization (NMF) has a good performance in solving sparsity problem. Manifold learning algorithms can identify and preserve the intrinsic geometrical structure of data. In order to get more accurate recommendation results, we propose a hybrid Slope One algorithm based on NMF. By constraining PNMF with graph regularization term, then we propose a weighted Slope One algorithm combined with neighborhood preserving PNMF. The hybrid algorithm has positive consequences for new data and can reduce computation complexity. Experimental show that optimized method has a good recommendation effect compared with tradition algorithm, it helps to solve the data sparsity problem and can improve the scalability.
- Badrul Sarwar, George Karypis, Joseph Konstan, and John Riedl. 2001. Item-based collaborative filtering recommendation algorithms. In Proceedings of the 10th international conference on World Wide Web (WWW '01). Association for Computing Machinery, New York, NY, USA, 285--295..Google Scholar
Digital Library
- Ricci, Francesco, et al. Recommender Systems Handbook. Springer-Verlag New York, Inc. 2011.Google Scholar
- D. Lemire, A. Maclachlan, Slope one predictors for online rating-based collaborative filtering, Proceedings of the 2005 SIAM International Conference on Data Mining, Society for Industrial and Applied Mathematics, 471--475, 2005.DOI: https://doi.org/10.1137/1.9781611972757.43Google Scholar
- Gianna M. Del Corso, Francesco Romani, Adaptive nonnegative matrix factorization and measure comparisons for recommender systems, Applied Mathematics and Computation, Volume 354, 2019, Pages 164--179, https://doi.org/10.1016/j.amc.2019.01.047.Google ScholarCross Ref
- Lee, D. Seung, H. Learning the parts of objects by nonnegative matrix factorization. Nature 401, 788--791 (1999)Google ScholarCross Ref
- https://sifter.org/~simon/journal/20061211.htmlGoogle Scholar
- Guo Jun-Peng, Hybrid Recommendation Algorithm based o Symbolic Data and Non-Negative Matrix Factorization [J], Journal of System & Management, 2015, 24(03): 372--378. (in Chinese with English Abstract)Google Scholar
- Donoho D, Stodden V. When does non-negative matrix factorization give a correct decomposition into parts? [C]// ELSEVIER, 2004. https://doi.org/10.1109/TII.2014.2308433Google Scholar
- Liang Hu, Yongheng Xing, Yanlei Gong, Kuo Zhao, Feng Wang, Nonnegative matrix tri-factorization with user similarity for clustering in point-of-interest, Neurocomputing, Volume 363, 2019, Pages 58--65, ISSN 0925-2312, https://doi.org/10.1016/j.neucom.2019.07.040.Google ScholarDigital Library
- B. Sarwar, G. Karypis, J. Konstan, and J. Riedl, "Application of dimensionality reduction in recommender systems---A case study," in Proc. ACM WebKDD, Boston, MA, USA, 2000, pp. 285--295.Google Scholar
- Cai, Deng, et al. "Locality preserving nonnegative matrix factorization." International Joint Conference on Artificial Intelligence Morgan Kaufmann Publishers Inc. 2009. DOI: https://doi.org/10.1109/SIBGRAPI.2009.48Google Scholar
- Belkin, Mikhail, and P. Niyogi. "Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering." Advances in neural information processing systems 14.6(2002): 585--591.Google Scholar
- G. Adomavicius and A. Tuzhilin, "Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions," in IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 6, pp. 734--749, June 2005. DOI: https://doi.org/10.1109/TKDE.2005.99Google ScholarDigital Library
- Mohit Sharma and George Karypis. 2019. Adaptive matrix completion for the users and the items in tail. In The World Wide Web Conference (WWW '19). Association for Computing Machinery, New York, NY, USA, 3223--3229. DOI: https://doi.org/10.1145/3308558.3313736Google ScholarDigital Library
- Koren, Y, R. Bell, and C. Volinsky. "Matrix Factorization Techniques for Recommender System." Computer 42.8(2009): 30--37Google Scholar
- Kim J, Park H. Fast Nonnegative Matrix Factorization: An Active-Set-Like Method and Comparisons[J]. SIAM Journal on Scientific Computing 33.6(2011): 3261--3281.DOI: https://doi.org/10.1137/110821172Google ScholarDigital Library
- HUANG Shu-qin, XU Yong, User similarity calculation method based on probabilistic matrix factorization and its recommendation application [J]. Journal of Shandong University (Natural Science), 2017, 52(11): 37--43+48.Google Scholar
- Yuan Z, Oja E. Projective Nonnegative Matrix Factorization for Image Compression and Feature Extraction[C]// Image Analysis, Scandinavian Conference, Scia, Joensuu, Finland, June. Springer-Verlag, 2005.Google Scholar
- Z. Yang and E. Oja, "Linear and Nonlinear Projective Nonnegative Matrix Factorization," in IEEE Transactions on Neural Networks, vol. 21, no. 5, pp. 734--749, May 2010. DOI: https://doi.org/10.1109/TNN.2010.2041361Google ScholarDigital Library
- X. Liu, S. Yan and H. Jin, "Projective Nonnegative Graph Embedding," in IEEE Transactions on Image Processing, vol. 19, no. 5, pp. 1126--1137, May 2010. DOI: https://doi.org/10.1109/TIP.2009.2039050Google ScholarDigital Library
- D. Cai, X. He, J. Han and T. S. Huang, "Graph Regularized Nonnegative Matrix Factorization for Data Representation," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 8, pp. 1548--1560, Aug. 2011. DOI: https://doi.org/10.1109/TPAMI.2010.231Google ScholarDigital Library
- Roweis, S. T. Nonlinear Dimensionality Reduction by Locally Linear Embedding[J]. Science, 290(5500): 2323--2326.Google Scholar
- Ju Bin, Qian Yun-tao, Ye Min-chao. Collaborative filtering algorithm based on structed projective nonnegative matrix factorization [J]. Journal of Zhejiang University (Engineering Science;, 2015, 49(07): 1319--1325. (in Chinese with English Abstract)Google Scholar
- Xiaofei He, Deng Cai, Shuicheng Yan and Hong-Jiang Zhang, "Neighborhood preserving embedding," Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1, Beijing, 2005, pp. 1208--1213 Vol. 2. DOI: https://doi.org/10.1109/ICCV.2005.167Google Scholar
- J. Wen, Z. Tian, X. Liu and W. Lin, "Neighborhood Preserving Orthogonal PNMF Feature Extraction for Hyperspectral Image Classification," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 6, no. 2, pp. 759--768, April 2013.Google ScholarCross Ref
- Y. Wang and Y. Zhang, "Nonnegative Matrix Factorization: A Comprehensive Review," in IEEE Transactions on Knowledge and Data Engineering, vol. 25, no. 6, pp. 1336--1353, June 2013. DOI: https://doi.org/10.1109/TKDE.2012.51Google ScholarDigital Library
- DONG Li-yan, FANG Yuan-cheng, Slope One algorithm based on nonnegative matrix factorization[J]. Journal of Zhejiang University (Engineering Science) 2019, 53(07): 1349--1353+1362. (in Chinese with English Abstract)Google Scholar
- Zhenyue Zhang, Keke Zhao, Hongyuan Zha, Inducible regularization for low-rank matrix factorizations for collaborative filtering, Neurocomputing, Volume 97, 2012, pp. 52--62, https://doi.org/10.1016/j.neucom.2012.05.010.Google ScholarDigital Library
- G. Ye and X. Zhao, "Improved SVD algorithm based on Slope One," 2018 Chinese Control and Decision Conference (CCDC), Shenyang, 2018, pp. 1002--1006.Google Scholar
Index Terms
- A Hybrid Slope One Collaborative Filtering Algorithm Based on Nonnegative Matrix Factorization
Recommendations
Collaborative filtering using non-negative matrix factorisation
Collaborative filtering is a popular strategy in recommender systems area. This approach gathers users' ratings and then predicts what users will rate based on their similarity to other users. However, most of the collaborative filtering methods have ...
Collaborative filtering using orthogonal nonnegative matrix tri-factorization
Collaborative filtering aims at predicting a test user's ratings for new items by integrating other like-minded users' rating information. The key assumption is that users sharing the same ratings on past items tend to agree on new items. Traditional ...
Combining review-based collaborative filtering and matrix factorization: A solution to rating's sparsity problem
AbstractAn important factor affecting the performance of collaborative filtering for recommendation systems is the sparsity of the rating matrix caused by insufficient rating data. Improving the recommendation model and introducing side ...
Highlights- Collaborative filtering suffers from the sparsity issue.
- The proposed method ...
Comments