《计算机应用研究》|Application Research of Computers

基于熵优化近邻选择的协同过滤推荐算法

Collaborative filtering recommendation algorithm based on entropy optimization nearest-neighbor selection

免费全文下载 (已被下载 次)  
获取PDF全文
作者 于阳,于洪涛,黄瑞阳
机构 国家数字交换系统工程技术研究中心,郑州 450002
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2017)09-2618-06
DOI 10.3969/j.issn.1001-3695.2017.09.012
摘要 协同过滤推荐系统的近邻选择环节中不仅没有考虑目标项目对用户间相似性计算的影响,而且也未考虑邻居用户对目标用户的推荐贡献能力,导致既降低了相似性计算的准确性,也提高了近邻集合中伪近邻的比例。针对这些问题,提出了一种基于熵优化近邻选择的协同过滤推荐算法。算法使用巴氏系数计算项目间相似性,并以此为权重加权计算用户间相似性。引入熵描述用户评分分布特性,根据评分分布差异性衡量邻居用户的推荐贡献能力。最后,利用双重准则共同计算推荐权重,并构建近邻集合。实验结果表明,该算法能够在不牺牲时间复杂度的条件下准确地选取近邻集合,提升推荐准确度。
关键词 协同过滤;近邻选择;相似性;巴氏系数;熵;推荐权重
基金项目 国家自然科学基金创新群体项目(61521003)
国家自然科学基金资助项目(61171108)
国家科技支撑计划资助项目(2014BAH30B01)
国家“973”计划资助项目(2012CB315901,2012CB315905)
本文URL http://www.arocmag.com/article/01-2017-09-012.html
英文标题 Collaborative filtering recommendation algorithm based on entropy optimization nearest-neighbor selection
作者英文名 Yu Yang, Yu Hongtao, Huang Ruiyang
机构英文名 ChinaNationalDigitalSwitchingSystemEngineering&TechnologicalR&DCenter,Zhengzhou450002,China
英文摘要 In the collaborative filtering recommendation system, the nearest-neighbor selection not only did not take into account the impact of the objective item on computing the similarity between pairs of users, but also did not consider the contribution of the neighbor users for the objective user, resulting in not only reducing the accuracy of similarity calculation, but also increasing the proportion of false nearest neighbors in the nearest-neighbor set. To address these problems, this paper put forward a collaborative filtering recommendation algorithm based on entropy optimization nearest-neighbor selection. Firstly, this algorithm calculated the similarity between items by using bhattacharyya coefficient, and it weighted the similarity between users by the similarity between the item and the target item. Secondly, it introduced the entropy to describe the distribution characteristics of the user’s rating, and measured the contribution of neighbor recommendation by the difference of the rating distribution. In the end, it used the double criteria to calculate the neighbor recommendation weights, and built the nearest neighbor set. Experimental results show that the proposed algorithm can accurately select the nearest-neighbor set to improve the degree of accuracy without sacrificing time complexity.
英文关键词 collaborative filtering; nearest-neighbor selection; similarity; Bhattacharyya coefficient; entropy; recommendation weight
参考文献 查看稿件参考文献
  [1] Chen Li, Chen Guanliang, Wang Feng. Recommender systems based on user reviews:the state of the art[J] . User Modeling and User-Adapted Interaction, 2015, 25(2):99-154.
[2] Mettouris C, Papadopoulos G A. Ubiquitous recommender systems[J] . Computing, 2014, 96(3):223-257.
[3] Jannach D, Zanker M, Felfernig A, et al. 推荐系统[M] . 蒋凡, 译. 北京:人民邮电出版社, 2013.
[4] Koren Y. Factorization meets the neighborhood:a multifaceted colla-borative filtering model[C] //Proc of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York:ACM Press, 2008:426-434.
[5] 熊忠阳, 刘芹, 张玉芳, 等. 基于项目分类的协同过滤改进算法[J] . 计算机应用研究, 2012, 29(2):493-496.
[6] Wang Jing, Ke Liangwen. Feature subspace transfer for collaborative filtering[J] . Neurocomputing, 2014, 136:1-6.
[7] Wang Weiqing, Chen Zhenyu, Liu Jia, et al. User-based collaborative filtering on cross domain by tag transfer learning[C] //Proc of the 1st International Workshop on Cross Domain Knowledge Discovery in Web and Social Network Mining. New York:ACM Press, 2012:10-17.
[8] 吴湖, 王永吉, 王哲, 等. 两阶段联合聚类协同过滤算法[J] . 软件学报, 2010, 21(5):1042-1054.
[9] Deng Shuiguang, Huang Longtao, Wu Jian, et al. Trust-based personalized service recommendation:a network perspective[J] . Journal of Computer Science & Technology, 2014, 29(1):69-80.
[10] Herlocker J, Konstan J A, Riedl J. An empirical analysis of design choices in neighborhood-based collaborative filtering algorithms[J] . Information Retrieval, 2002, 5(4):287-310.
[11] Kim T H, Yang S B. An effective threshold-based neighbor selection in collaborative filtering[C] //Proc of the 29th European Conference on IR Research. [S. l. ] :Springer-Verlag, 2007:712-715.
[12] Baltrunas L, Ricci F. Locally adaptive neighborhood selection for collaborative filtering recommendations[C] //Adaptive Hypermedia and Adaptive Web-Based Systems. Berlin:Springer, 2008:22-31.
[13] Zhang Liang, Xiao Bo, Guo Jun. An approach of selecting right neighbors for collaborative filtering[C] //Proc of the 4th International Conference on Innovative Computing, Information and Control. 2009:1057-1060.
[14] Anand D, Bharadwaj K K. Utilizing various sparsity measures for enhancing accuracy of collaborative recommender systems based on local and global similarities[J] . Expert Systems with Applications, 2011, 38(5):5101-5109.
[15] 贾冬艳, 张付志. 基于双重邻居选取策略的协同过滤推荐算法[J] . 计算机研究与发展, 2013, 50(5):1076-1084.
[16] Kaleli C. An entropy-based neighbor selection approach for collaborative filtering[J] . Knowledge-Based Systems, 2014, 56:273-280.
[17] Park Y, Park S, Jung W, et al. Reversed CF:a fast collaborative filtering algorithm using a k -nearest neighbor graph[J] . Expert Systems with Applications, 2015, 42(8):4022-4028.
[18] 张佳, 林耀进, 林梦雷, 等. 基于目标用户近邻修正的协同过滤算法[J] . 模式识别与人工智能, 2015, 28(9):802-810.
[19] Ning Xia, Desrosiers C, Karypis G. A comprehensive survey of neighborhood-based recommendation methods[M] //Recommender Systems Handbook. 2015:37-76.
[20] Nielsen F, Boltz S. The burbea-rao and bhattacharyya centroids[J] . IEEE Trans on Information Theory, 2011, 57(8):5455-5466.
[21] Patra B K, Launonen R, Ollikainen V, et al. A new similarity mea-sure using Bhattacharyya coefficient for collaborative filtering in sparse data[J] . Knowledge-Based Systems, 2015, 82(2):163-177.
[22] 朱郁筱, 吕琳媛. 推荐系统评价指标综述[J] . 电子科技大学学报, 2012, 41(2):163-175.
收稿日期 2016/6/19
修回日期 2016/7/25
页码 2618-2623
中图分类号 TP181
文献标志码 A