›› 2012,Vol. 27 ›› Issue (3): 567-576.doi: 10.1007/s11390-012-1244-x

• • 上一篇    下一篇

基于指数模糊C均值的协同过滤

Kiatichai Treerattanapitak and Chuleerat Jaruskulchai   

  • 收稿日期:2011-09-01 修回日期:2012-03-01 出版日期:2012-05-05 发布日期:2012-05-05

Exponential Fuzzy C-Means for Collaborative Filtering

Kiatichai Treerattanapitak and Chuleerat Jaruskulchai   

  1. Department of Computer Science, Kasetsart University, 50 Ngamwongwan Rd., Jatuchak, Bangkok, Thailand
  • Received:2011-09-01 Revised:2012-03-01 Online:2012-05-05 Published:2012-05-05
  • About author:Kiatichai Treerattanapitak is currently a Ph.D. candidate in Kasetsart University, Thailand. He is also a senior consultant of envi-ronmental resources management in Thailand. His research interests in-clude artificial inteligence, data min-ing, and clustering.

协同过滤(CF)是成功应用于推荐系统中的最为流行的技术之一.通过从以往的用户那里收集信息,,协同过滤可以预测出具有相同意见的当前用户的兴趣.在协同过滤研究领域使用最广泛的方法是矩阵分解,如奇异值分解(SVD).但是,很多有名的推荐系统并没有采用这类方法,而是仍然坚持使用近邻模型,原因在于近邻模型的简单性和可解释性.但近邻模型中存在的很多问题限制了它获得更高的预测精度.为了解决这些问题,我们提出了一种新的指数模糊聚类(XFCM)算法,通过一个指数方程来重新描述聚类目标函数,以获得更好的隶属度指派.该算法通过积极排除一些不相关的数据,将数据指派到各个簇中,因此优于其他模糊C均值(FCM)算法.实验结果表明,基于指数模糊C均值聚类算法的协同过滤方法在100k和1M 的MovieLens数据集上的绝对平均误差比基于项目的方法改善了6.9%,比SVD方法改善了3.0%.

Abstract: Collaborative filtering (CF) is one of the most popular techniques behind the success of recommendation system. It predicts the interest of users by collecting information from past users who have the same opinions. The most popular approaches used in CF research area are Matrix factorization methods such as SVD. However, many well-known recommendation systems do not use this method but still stick with Neighborhood models because of simplicity and explainability. There are some concerns that limit neighborhood models to achieve higher prediction accuracy. To address these concerns, we propose a new exponential fuzzy clustering (XFCM) algorithm by reformulating the clustering's objective function with an exponential equation in order to improve the method for membership assignment. The proposed method assigns data to the clusters by aggressively excluding irrelevant data, which is better than other fuzzy C-means (FCM) variants. The experiments show that XFCM-based CF improved 6.9% over item-based method and 3.0% over SVD in terms of mean absolute error for 100K and 1M MovieLens dataset.

[1] Herlocker J L, Konstan J A, Borchers A, Riedl J. An algorith-mic framework for performing collaborative filtering. In Proc.the 22nd ACM SIGIR Conf. Research and Development inInformation Retrieval, Aug. 1999, pp.230-237.

[2] Sarwar B, Karypis G, Konstan J, Riedl, J. Item-based col-laborative filtering recommendation algorithms. In Proc. the10th Int. Conf. World Wide Web, May 2001, pp.285-295.

[3] Karypis G. Evaluation of item-based top-N recommendationalgorithms. In Proc. the 10th Conf. Information and Know-ledge Management, Nov. 2001, pp.247-254.

[4] Linden G, Smith B, York J. Amazon.com recommendations:Item-to-item collaborative filtering. IEEE Internet Comput-ing, 2003, 7(1): 76-80.

[5] Bell R, Koren Y. Scalable collaborative filtering with jointlyderived neighborhood interpolation weights. In Proc. the 7thInt. Conf. Data Mining, Oct. 2007, pp.43-52.

[6] Koren Y, Bell R. Advanced in collaborative filtering. In Reco-mmender Systems Handbook (1st edition), Springer, 2011, pp.145-186.

[7] Sarwar B M, Karypis G, Konstan J A, Riedl J T. Applica-tion of dimensionality reduction in recommender system —— Acase study. In ACM WebKDD Web Mining for ECommenceWorkshop, Aug. 2000.

[8] Vozalis M, Markos A, Margaritis K G. Evaluation of standardSVD-based techniques for collaborative filtering. In Proc. the9th Hellenic European Research on Computer Mathematicsand its Applications, Sept. 2009.

[9] Rendle S. Factorization machines. In Proc. the 10th Int.Conf. Data Mining, Dec. 2010, pp.995-1000.

[10] Ali K, van Stam W. TiVo: Making show recommendations us-ing a distributed collaborative filtering architecture. In Proc.the 10th ACM SIGKDD Int. Conf. Knowledge Discoveryand Data Mining, Aug. 2004, pp.394-401.

[11] Koren Y. Factorization meets the neighborhood: A multi-faceted collaborative filtering model. In Proc. the 14th ACMSIGKDD Int. Conf. Knowledge Discovery and Data Mining,Aug. 2008, pp.426-434.

[12] Liu N N, Yang Q. EigenRank: A ranking-oriented approach tocollaborative filtering. In Proc. the 31st Conf. ACM SIGIRon Information Retrieval, Jul. 2008, pp.83-90.

[13] Treerattnapitak K, Jaruskulchai C. Entropy based fuzzy C-mean for item-based collaborative filtering. In Proc. the 9thInt. Symposium on Communication and Information Tech-nology, Sept. 2009, pp.881-886.

[14] Treerattnapitak K, Jaruskulchai C. Items based fuzzy C-meanclustering for collaborative filtering. Information TechnologyJournal, 2009, 5(10): 30-34.

[15] Jin R, Si L. A study of methods for normalizing user ratingsin collaborative filtering. In Proc. the 27th Conf. ACM SI-GIR on Research and Development in Information Retrieval,Jul. 2004, pp.568-569.

[16] Breese J S, Heckerman D, Kadie C. Empirical analysis of pre-dictive algorithms for collaborative filtering. In Proc. the14th Conf. Uncertainty in Artificial Intelligence, Jul. 1998,pp.43-52.

[17] George T, Merugu S. A scalable collaborative filtering frame-work based on co-clustering. In Proc. the 5th IEEE Int.Conf. Data Mining, Nov. 2005, pp.625-628.

[18] Ungar L H, Foster D P. Clustering methods for collabora-tive filtering. In Proc. AAAI Workshop on RecommendationSystem, Jul. 1998.

[19] Pham M C, Cao Y, Klamma R, Jarke M. A clustering ap-proach for collaborative filtering recommendation using socialnetwork analysis. Journal of Universal Computer Science,2011, 17(4): 583-604.

[20] Gong S. A collaborative filtering recommendation algorithmbased on user clustering and item clustering. Journal of Soft-ware, 2010, 5(7): 745-752.

[21] Wu J, Li T. A modified fuzzy C-means algorithm for collabo-rative filtering. In Proc. the 2nd KDD Workshop on Large-Scale Recommender Systems and the Netflix Prize Competi-tion, Aug. 2008, Article No. 2.

[22] Treerattnapitak K, Jaruskulchai C. Membership enhancementwith exponential fuzzy clustering for collaborative filtering.In Proc. the 17th Int. Conf. Neural Information Processing,Nov. 2010, pp.559-566.

[23] Wang H, Pei J. Clustering by pattern similarity. Journal ofComputer Science and Technology, 2008, 23(4): 481-496.

[24] Wattanachon U, Suksawatchon J, Lursinsap C. Nonlineardata analysis using a new hybrid data clustering algorithm.In Lecture Notes in Computer Science 5476, TheeramunkongT et al. (eds.), Springer-Verlag, 2009, pp.160-171.

[25] Dunn J C. A fuzzy relative of the ISODATA process and itsuse in detecting compact well-separated clusters. Journal ofCybernetics., 1973, 3(3): 32-57.

[26] Bezdek J C. Pattern Recognition with Fuzzy Objective Func-tion Algoritms. New York: Plenum Press, 1981.

[27] Miyamoto S, Mukaidono M. Fuzzy C-means as a regulariza-tion and maximum entropy approach. In Proc. the 7th Int.Fuzzy System Association World Congress (IFSA 1997), Jun.1997, 2: 86-92.

[28] Miyamoto S, Ichihashi H, Katsuhiro H. Algorithms for FuzzyClustering: Methods in c-Means Clustering with Applica-tions. Springer-Verlag Berlin Heidelberg, 2008.

[29] Goharian N, El-Ghazawi T A, Grossman D A, Chowdhury A.On the enhancements of a sparse matrix information retrievalapproach. In Proc. the Int. Conf. Parallel and DistributedProcessing Technology and Application, Jun. 2000.
No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 姚荣; 康泰; 陈廷槐;. Algorithms for the Determination of Cutsets in a Hypergraph[J]. , 1990, 5(1): 41 -46 .
[2] 朱明远;. Two Congruent Semantics for Prolog with CUT[J]. , 1990, 5(1): 82 -91 .
[3] 韩建超; 史忠植;. Formalizing Default Reasoning[J]. , 1990, 5(4): 374 -378 .
[4] 蔡士杰; 张福炎;. A Fast Algorithm for Polygon Operations[J]. , 1991, 6(1): 91 -96 .
[5] 王能斌; 刘海青;. An Intelligent Tool to Support Requirements Analysis and Conceptual Design of Database Design[J]. , 1991, 6(2): 153 -160 .
[6] 林珊;. Using a Student Model to Improve Explanation in an ITS[J]. , 1992, 7(1): 92 -96 .
[7] 费翔林; 廖雷; 王和珍; 汪承藻;. Structured Development Environment Based on the Object-Oriented Concepts[J]. , 1992, 7(3): 193 -201 .
[8] 周勇; 唐泽圣;. Constructing Isosurfaces from 3D Data Sets Taking Account of Depth Sorting of Polyhedra[J]. , 1994, 9(2): 117 -127 .
[9] 王晖; 刘大有; 王亚飞;. Sequential Back-Propagation[J]. , 1994, 9(3): 252 -260 .
[10] 刘健; 陈治明; 钟彦儒; 杜忠;. Compact DC-DC Converter for Pocket Micro-Controller Systems[J]. , 1996, 11(6): 607 -614 .
版权所有 © 《计算机科学技术学报》编辑部
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn
总访问量: