融合显/隐式反馈的协同排序算法

doi:10.11772/j.issn.1001-9081.2015.05.1328

计算机应用 ›› 2015, Vol. 35 ›› Issue (5): 1328-1332.DOI: 10.11772/j.issn.1001-9081.2015.05.1328

融合显/隐式反馈的协同排序算法

李改^1,2,3

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

收稿日期:2014-12-19 修回日期:2015-02-03 出版日期:2015-05-10 发布日期:2015-05-14
通讯作者: 李改
作者简介:李改(1981-),男,湖北松滋人,讲师,博士研究生,CCF会员,主要研究方向:机器学习、人工智能、推荐系统.
基金资助:
国家自然科学基金资助项目(61003140,61033010);2011年度高校基本科研业务费中山大学青年教师培育项目(理工科)(111gpy58);佛山市产学研专项资金项目(2012HC100303);广东省高等职业教育教学改革2014年度项目(201401294);广东省高等职业教育教学改革2013年度项目(20130201109);广东省高等职业教育教学改革2012年度项目(20120302050);顺德职业技术学院2013年度校级教学改革与研究重点项目(2013-SZJGXM13).

Collaborative ranking algorithm by explicit and implicit feedback fusion

LI Gai^1,2,3

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

Received:2014-12-19 Revised:2015-02-03 Online:2015-05-10 Published:2015-05-14

摘要/Abstract

摘要：

之前有关协同排序算法的研究没有充分利用数据集中信息的问题,要么只侧重于研究显式评分数据,要么只侧重于研究隐式评分数据,目前还没有人运用排序学习的思想把二者结合起来进行研究.针对之前研究的不足,在最新的扩展的少即是好协同过滤(xCLiMF)模型和最经典的变形的奇异值分解(SVD++)算法的基础上,提出了一种融合显/隐式反馈的协同排序算法MERR_SVD++来直接优化排序学习的评价指标ERR.在实际数据集上实验验证,与经典的xCLiMF、Cofi排序(CofiRank)、PopRec、Random算法相比,MERR_SVD++算法在归一化折损累积增益(NDCG)和预期的相关性排序(ERR)这两个评价指标下性能均提高了25.9%以上,而且算法运算时间与评分点个数线性相关.由于MERR_SVD++算法推荐精度高、可扩展性好,因此适用于处理大数据,在互联网信息推荐领域具有广泛的应用前景.

关键词: 推荐系统, 协同过滤, 协同排序, 隐式反馈, 显式反馈

Abstract:

The problem of the previous research about collaborative ranking is that it does not make full use of the information in the dataset, either focusing on explicit feedback data, or focusing on implicit feedback data. Until now, nobody researches collaborative ranking algorithm by explicit and implicit feedback fusion. In order to overcome the defects of prior research, a new collaborative ranking algorithm by explicit and implicit feedback fusion namedMERR_SVD++ was proposed to optimize Expected Reciprocal Rank (ERR) based on the newest Extended Collaborative Less-is-More Filtering (xCLiMF) model and Singular Value Decomposition++ (SVD++) algorithm. The experimental results on practical datasets show that, the values of Normalized Discounted Cumulative Gain (NDCG) and ERR for MERR_SVD++ are increased by 25.9% compared with xCLiMF, Cofi Ranking (CofiRank), PopRec and Random collaborative ranking algorithms, and the running time of MERR_SVD++ showed a linear correlation with the number of ratings. Because of the high precision and the good expansibility, MERR_SVD++ is suitable for processing big data, and has wide application prospect in the field of Internet information recommendation.

Key words: recommended system, collaborative filtering, collaborative ranking, implicit feedback, explicit feedback

中图分类号:

李改. 融合显/隐式反馈的协同排序算法[J]. 计算机应用, 2015, 35(5): 1328-1332.

LI Gai. Collaborative ranking algorithm by explicit and implicit feedback fusion[J]. Journal of Computer Applications, 2015, 35(5): 1328-1332.

参考文献

[1] ADOMAVICIUS G, TUZHILIN A. Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions[J]. IEEE Transactions on Knowledge and Data Engineering, 2005,17(6):734-749.
[2] PAN R, ZHOU Y, CAO B,et al. One-class collaborative filtering [C]// Proceedings of the 22nd International Conference on Data Mining. Piscataway: IEEE, 2008:502-511.
[3] HERNANDEZ-LOBOTA J M, HOULSBY N, GHAHRAMANI Z B. Probabilistic matrix factorization with non-random missing data[C]// Proceedings of the 31nd International Conference on Machine Learning. Piscataway: IEEE, 2014: 1257-1264.
[4] RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: Bayesian personalized ranking from implicit feedback[C]// Proceedings of the 22nd International Conference on Uncertainty in Artificial Intelligence. Montreal: AUAI Press, 2009: 52-461.
[5] LIU T Y. Learning to rank for information retrieval[M]. New York: Springer, 2011: 1-304.
[6] SUHRID B, SUMIT C. Collaborative ranking[C]// Proceedings of the 2012 ACM International Conference on Web Search and Data Mining. New York: ACM, 2012:143-152.
[7] KOREN Y. Factorization meets the neighborhood: a multifaceted collaborative filtering model[C]// Proceedings of the 25th International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2008: 426-434.
[8] SHI Y, KARATZOGLOU A, BALTRUNAS L, et al. xCLiMF: Optimizing expected reciprocal rank for data with multiple levels of relevance[C]// Proceedings of the 6th ACM Conference on Recommender Systems. New York: ACM, 2013: 431-433.
[9] PAN W, LI C. GBPR: group preference based Bayesian personalized ranking for one-class collaborative filtering[C]// Proceedings of the 26th International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2009: 667-676.
[10] SHI Y, KARATZOGLOU A, BALTRUNAS L, et al. CLiMF: Collaborative Less-is-More Filtering [C]// Proceedings of the 23rd International Conference on Artificial Intelligence. New York: ACM, 2013: 3077-3081.
[11] SREBRO N, JASON D M, TOMMI J. Maximum-margin matrix factorization[C]// Proceedings of the 22nd International Conference on Advances in Neural Information Processing Systems. Cambridge: MIT Press, 2005: 252-260.
[12] WEIMER M, KARATZOGLOU A, LE Q, et al. CofiRank-maximum margin matrix factorization for collaborative ranking [C]// Proceedings of the 22nd International Conference on Advances in Neural Information Processing Systems. Cambridge: MIT Press, 2007: 1593-1600.
[13] LIU N, ZHAO M, YANG Q. Probabilistic latent preference analysis for collaborative filtering [C]// Proceedings of the 2009 ACM International Conference on Information and Knowledge Management. New York: ACM, 2009: 759-766.
[14] LIU N, YANG Q. EigenRank: a ranking-oriented approach to collaborative filtering[C]// Proceedings of the 22nd International Conference on Research and Development in Information Retrieval. New York: ACM, 2008: 83-90.
[15] SHI Y, KARATZOGLOU A, BALTRUNAS L, et al. CLiMF: learning to maximize reciprocal rank with collaborative less-is-more filtering[C]// Proceedings of the 5th ACM Conference on Recommender Systems. New York: ACM, 2012: 139-146.

融合显/隐式反馈的协同排序算法

Collaborative ranking algorithm by explicit and implicit feedback fusion

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	包玄, 陈红梅, 肖清. 融入时间的兴趣点协同推荐算法[J]. 计算机应用, 2021, 41(8): 2406-2411.
[2]	刘欢, 李晓戈, 胡立坤, 胡飞雄, 王鹏华. 基于知识图谱驱动的图神经网络推荐模型[J]. 计算机应用, 2021, 41(7): 1865-1870.
[3]	林怿星, 唐华. 基于异构信息网络的混合推荐模型[J]. 计算机应用, 2021, 41(5): 1348-1355.
[4]	高铭蔚, 桑楠, 杨茂林. 基于胶囊网络的交互式网络电视视频点播推荐模型[J]. 《计算机应用》唯一官方网站, 2021, 41(11): 3171-3177.
[5]	任柯舟, 彭甫镕, 郭鑫, 王喆, 张晓静. 动态融合社交信息的社会化推荐[J]. 计算机应用, 2021, 41(10): 2806-2812.
[6]	李翔锟, 贾彩燕. 融合重叠社区正则化及隐式反馈的协同过滤方法[J]. 计算机应用, 2021, 41(1): 53-59.
[7]	李春英, 汤庸, 肖政宏, 李天送. 学术社交网络中的权威学者推荐模型[J]. 计算机应用, 2020, 40(9): 2594-2599.
[8]	田保军, 刘爽, 房建东. 融合主题信息和卷积神经网络的混合推荐算法[J]. 计算机应用, 2020, 40(7): 1901-1907.
[9]	陈曦, 梅广, 张金金, 许维胜. 融合知识图谱和协同过滤的学生成绩预测方法[J]. 计算机应用, 2020, 40(2): 595-601.
[10]	秦婧, 张青博, 王斌. 关注长尾物品的推荐方法[J]. 计算机应用, 2020, 40(2): 454-458.
[11]	张文龙, 钱付兰, 陈洁, 赵姝, 张燕平. 基于双重最相关注意力网络的协同过滤推荐算法[J]. 计算机应用, 2020, 40(12): 3445-3450.
[12]	樊玮, 谢聪, 肖春景, 曹淑燕. 基于组合类别空间的随机游走推荐算法[J]. 计算机应用, 2019, 39(4): 984-988.
[13]	雷曼, 龚琴, 王纪超, 王保群. 基于标签权重的协同过滤推荐算法[J]. 计算机应用, 2019, 39(3): 634-638.
[14]	刘彤, 曾诚, 何鹏. 基于用户网络嵌入的民宿房源推荐方法[J]. 计算机应用, 2019, 39(11): 3398-3402.
[15]	许朝, 孟凡荣, 袁冠, 李月娥, 刘肖. 融合地点影响力的兴趣点推荐算法[J]. 计算机应用, 2019, 39(11): 3178-3183.