鲁棒的单类协同排序算法

李改; 李磊

doi:10.16383/j.aas.2015.c140231

鲁棒的单类协同排序算法

doi: 10.16383/j.aas.2015.c140231

李改^1,2,3, ,,
李磊^1,3

1.
中山大学信息科学与技术学院广州 510006;
2.
顺德职业技术学院电子与信息工程学院顺德 528333;
3.
中山大学软件研究所广州 510275

基金项目:

国家自然科学基金(61003140,61033010),中山大学高性能与网格计算平台资助

详细信息

作者简介:
李磊博士. 中山大学信息科学与技术学院教授. 主要研究方向为数据库, 数据挖掘, 人工智能.E-mail: 21115903@qq.com

通讯作者:
李改中山大学信息科学与技术学院博士研究生. 2005 年获得中山大学信息科学与技术学院硕士学位. 主要研究方向为推荐系统, 数据挖掘. 本文通信作者. E-mail: ligai999@126.com

计量
- 文章访问数: 1665
- HTML全文浏览量: 43
- PDF下载量: 2054
- 被引次数: 0
出版历程
- 收稿日期: 2014-04-08
- 修回日期: 2014-09-12
- 刊出日期: 2015-02-20

Robust Ranking Algorithms for One-class Collaborative Filtering

LI Gai^{1,2,3
, ,},
LI Lei^1,3

1.
School of Information Science and Technology, Sun Yat-sen University, Guangzhou 510006;
2.
School of Electronics and Information Engineering, Shunde Polytechnic, Shunde 528333;
3.
Software Institute, Sun Yat-sen University, Guangzhou 510275

Funds:

Supported by National Natural Science Foundation of China (61003140, 61033010), and High Performance and Grid Computing Platform of Sun Yat-sen University

摘要

摘要: 单类协同过滤(One-class collaborative filtering, OCCF)问题是当前的一大研究热点.之前的研究所提出的算法对噪声数据很敏感,因为训练数据中的噪声数据将给训练过程带来巨大影响,从而导致算法的不准确性.文中引入了Sigmoid成对损失函数和Fidelity成对损失函数,这两个函数具有很好的灵活性,能够和当前最流行的基于矩阵分解(Matrix factorization, MF)的协同过滤算法和基于最近邻(K-nearest neighbor, KNN)的协同过滤算法很好地融合在一起,进而提出了两个鲁棒的单类协同排序算法,解决了之前此类算法对噪声数据的敏感性问题.基于Bootstrap抽样的随机梯度下降法用于优化学习过程.在包含有大量噪声数据点的实际数据集上实验验证,本文提出的算法在各个评价指标下均优于当前最新的单类协同排序算法.
- 推荐系统 /
- 单类协同过滤 /
- 协同排序 /
- 隐式反馈 /
- 成对损失函数
Abstract: The problem of ranking for one-class collaborative filtering (OCCF) is a research focus. One drawback of the existing ranking algorithms for OCCF is noise sensitivity, because the noisy data of training data might bring big influences to the training process and lead to inaccuracy of the algorithm. In this paper, in order to solve the noise sensitivity problem of the ranking algorithms, we propose two robust ranking algorithms for OCCF by using the pairwise sigmoid/fidelity loss functions that are flexible and can be easily adopted by the popular matrix factorization (MF) model and the K-nearest-neighbor (KNN) model. We use stochastic gradient descent with bootstrap sampling to optimize the two robust ranking algorithms. Experimental results on three practical datasets containing a large number of noisy data show that our proposed algorithms outperform several state-of-the-art ranking algorithms for OCCF in terms of different evaluation metrics.
- Recommender systems /
- one-class collaborative flltering (OCCF) /
- collaborative ranking /
- implicit feedback /
- pairwise loss function

HTML全文

参考文献(27)

[1]	Adomavicius G, Tuzhilin A. Toward the next generationof recommendersystems: A survey of the state-of-the-art and possible extensions. IEEE Transactions on Knowledge and Data Engineering, 2005, 17(6): 734-749
[2]	[2] Pan R, Zhou Y H, Cao B, Liu N N, Lukose R, Scholz M, Yang Q. One-class collaborative filtering. In: Proceedings of the 2008 IEEE International Conference on Data Mining. Pisa, Italy: IEEE, 2008. 502-511
[3]	[3] Pan R, Scholz M. Mind the gaps: weighting the unknown in large-scale one-class collaborative filtering. In: Proceedings of the 22nd International Conference on Knowledge Discovery and Data Mining. Paris, France: ACM, 2009. 667-676
[4]	[4] Hu Y, Koren Y, Volinsky C. Collaborative filtering for implicit feedback datasets. In: Proceedings of the 8th IEEE International Conference on Data Mining. Pisa, Italy: IEEE, 2008. 263-272
[5]	[5] Sindhwani V, Bucak S S, Hu J, Mojsilovi A. One-class matrix completion with low-density factorizations. In: Proceedings of the 10th IEEE International Conference on Data Mining. Sydney, NSW: IEEE, 2010. 1055-1060
[6]	[6] Rendle S, Freudenthaler C, Gantner Z, Schmidt-Thieme L. BPR: Bayesian personalized ranking from implicit feedback. In: Proceedings of the 25th International Conference on Uncertainty in Artificial Intelligence. Montreal, Canada: IEEE, 2009. 452-461
[7]	[7] Shi Y, Karatzoglou A, Baltrunas L. CLiMF: collaborative less-is-more filtering. In: Proceedings of the 23rd International Conference on Artificial Intelligence. Beijing, China: ACM, 2013. 3077-3081
[8]	[8] Shi Y, Karatzoglou A, Baltrunas L, Larson M, Oliver N, Hanjalic A. CLiMF: learning to maximize reciprocal rank with collaborative less-is-more filtering. In: Proceedings of the 6th ACM Conference on Recommender Systems. Dublin. Irland: ACM, 2012. 139-146
[9]	[9] Pan W K, Chen L. GBPR: group preference based Bayesian personalized ranking for one-class collaborative filtering. In: Proceedings of the 23rd International Conference on Artificial Intelligence. Beijing, China: ACM, 2013. 3007-3011
[10]	Du L, Li X, Shen Y D. User graph regularized pairwise matrix factorization for item recommendation. In: Proceedings of the 7th International Conference on Advanced Data Mining and Applications. Berlin, German: ACM, 2011. 372-385
[11]	Rendle S, Schmidt-Thieme L. Pairwise interaction tensor factorization for personalized tag recommendation. In: Proceedings of the 3rd ACM International Conference on Web Search and Data Mining. New York, USA: ACM, 2010. 81-90
[12]	Yang S, Long B, Alexander J, Zha H, Zheng Z. Collaborative competitive filtering: learning recommender using context of user choice. In: Proceedings of the 34th ACM International Conference on Research and Development in Information Retrieval. Beijing, China: ACM, 2011. 295-304
[13]	Kanagal B, Ahmed A, Pandey S, Josifovski V, Yuan J, Garcia-Pueyo L. Supercharging recommender systems using taxonomies for learning user purchase behavior. In: Proceedings of the 2012 VLDB Endowment. Istanbul, Turkey: ACM, 2012. 956-967
[14]	Carvalho V, Elsas J, Cohen W, Carbonell J. A meta-learning approach for robust rank learning. In: Proceedings of the 22nd International Conference on Research and Development in Information Retrieval. Singapore, Singapore: ACM, 2008. 208-214
[15]	Tsai M F, Liu T Y, Qin T, Chen H H, Ma W Y. Frank: a ranking method with fidelity loss. In: Proceedings of the 21st International Conference on Research and Development in Information Retrieval. Amsterdam, Holland: ACM, 2007. 383-390
[16]	Scholkopf B, Platt J C, Shawe-Taylor J, Smola A J, Williamson R C. Estimating the support of a high-dimensional distribution. Neural Computation, 2001, 13(7): 1443-1471
[17]	Zhang S, Wang W H, Ford J, Makedon F, Pearlman J. Using singular value decomposition approximation for collaborative filtering. In: Proceedings of the 7th IEEE International Conference on E-Commerce. Mnchen, German: IEEE, 2005. 257-264
[18]	Li Y N, Zhai C X, Hu J, Chen Y. Improving one-class collaborative filtering by incorporating rich user information. In: Proceedings of the 19th ACM International Conference on Information and Knowledge Management. New York, USA: ACM, 2010. 959-968
[19]	Kaya H, Alpaslan F N. Using social networks to solve data sparsity problem in one-class collaborative filtering. In: Proceedings of the 7th IEEE International Conference on Information Technology. Las Vegas, NV: IEEE, 2010. 249-252
[20]	Wang C, Blei D M. Collaborative topic modeling for recommending scientific articles. In: Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Diego, CA: ACM, 2011. 448-456
[21]	Purushotham S, Liu Y, Kuo C. Collaborative topic regression with social matrix factorization for recommendation systems. In: Proceedings of the 29th ACM International Conference on Machine Learing. Edinburgh, Scotland, UK: ACM, 2012. 1255-1265
[22]	Ding X T, Jin X M, Li Y J, Li L H. Celebrity recommendation with collaborative social topic regression. In: Proceedings of the 23rd International Conference on Artificial Intelligence. Beijing, China: ACM, 2013. 2612-2618
[23]	Tang J, Yan J, Ji L, Zhang M, Guo S D, Liu N, Wang X F, Chen Z. Collaborative users' brand preference mining across multiple domains from implicit feedbacks. In: Proceedings of the 25th AAAI Conference on Artificial Intelligence. San Francisco, USA: AAAI, 2011. 477-482
[24]	Celma O. Music Recommendation and Discovery in the Long Tail. New York: Springer, 2010.
[25]	Liu Tie-Yan. Learning to Rank for Information Retrieval. New York: Springer, 2011.
[26]	Deshpande M, Karypis G. Item-based top-N recommendation algorithms. ACM Transactions on Information Systems, 2004, 22(1):143-177
[27]	Yang Zhen, Lai Ying-Xu, Duan Li-Juan, Li Yu-Jian, Xu Xin. Spam collaborative filtering in enron E-mail network. Acta Automatica Sinica, 2012, 38(3): 399-411 (杨震, 赖英旭, 段立娟, 李玉鑑, 许昕. 邮件网络协同过滤机制研究. 自动化学报, 2012, 38(3): 399-411)