您的位置:山东大学 -> 科技期刊社 -> 《山东大学学报(理学版)》

山东大学学报(理学版) ›› 2017, Vol. 52 ›› Issue (11): 37-43.doi: 10.6040/j.issn.1671-9352.0.2017.326

• • 上一篇    下一篇

基于概率矩阵分解的用户相似度计算方法及推荐应用

黄淑芹1,2,徐勇1,王平水1   

  1. 1.安徽财经大学管理科学与工程学院, 安徽 蚌埠 233030;2.中国科学技术大学计算机科学与技术学院, 安徽 合肥 230027
  • 收稿日期:2017-05-30 出版日期:2017-11-20 发布日期:2017-11-17
  • 作者简介:黄淑芹(1976— ),女,汉族,硕士,讲师,研究方向为数据挖掘和个性化推荐的研究. E-mail:393054429@qq.com
  • 基金资助:
    国家社会科学基金资助项目(15BTQ043,16BTQ085);安徽省自然科学基金资助项目(1408085MF127);安徽财经大学科研基金资助项目(ACKY1755)

User similarity calculation method based on probabilistic matrix factorization and its recommendation application

HUANG Shu-qin1, 2, XU Yong1, WANG Ping-shui1   

  1. 1. School of Management Science and Engineering, Anhui University of Finance and Economics, Bengbu 233030, Anhui, China;
    2. School of Computer Science and Technology, University of Science and Technology of China, Hefei 230027, Anhui, China
  • Received:2017-05-30 Online:2017-11-20 Published:2017-11-17

摘要: 用户相似度计算的合理性直接影响到协同过滤推荐的效果。提出了一种基于时序行为关系的用户消费网络图构建方法,通过定义用户间非对称相似度计算方法,确定用户间的初始相似度矩阵,然后利用概率矩阵分解的方法重构用户的相似度矩阵,挖掘潜在的用户近邻,将近邻关系应用到目标用户的项目推荐中,同时提出了基于时序行为关系和矩阵分解的协同过滤推荐框架结构。在实际数据集上对具体参数进行实验,并和其他方法进行了比较。实验结果表明,该方法可以有效提高协同过滤推荐效果。

关键词: 时序行为关系, 概率矩阵分解, 非对称用户相似度, 近邻

Abstract: The rationality of user similarity computation directly affects the effect of collaborative filtering. A method is proposed to construct a user consumption network and a calculation formula of asymmetric user similarity was defined. The initial user similarity matrix is set up based on the user asymmetric similarity values and reconstructed via probabilistic matrix factorization. And then, potential nearest neighbor mined is applied to personalized recommendation of the target user. Meanwhile, a collaborative filtering recommendation framework is put forward based on sequential behavior and probabilistic matrix factorization. The actual parameters are experimented on real-world dataset and compared with other methods. Experimental results show that the proposed method can improve the effect of collaborative filtering recommendation.

Key words: nearest neighbor, asymmetric user similarity, sequential behavior relationship, probabilistic matrix factorization

中图分类号: 

  • TP393
[1] EKSTRAND M D, RIEDL J T, KONSTAN J A. Collaborative filtering recommender systems[J]. Foundations and Trends in Human-Computer Interaction, 2010, 4(2): 81-173.
[2] LIU Qi, CHEN Enhong, XIONG Hui, et al. Enhancing collaborative filtering by user interests expansion via personalized ranking[J]. IEEE Transactions on Systems, Man and Cybernetics-part B: Cybernetics, 2012, 42(1):218-233
[3] CHOI K, SUH Y. A new similarity function for selecting neighbors for each target item in collaborative filtering[J]. Knowledge-Based Systems. 2013, 37:146-153
[4] SHI Y, LARSON M, HANJALIC A. Exploiting user similarity based on rated-item pools for improved user-based collaborative filtering[C] // Proceeding of the third ACM conference on Recommender systems(RecSys’09). New York: ACM, 2009: 125-132.
[5] BREESE J S, HECKERMAN D, KADIE C. Empirical analysis of predictive algorithms for collaborative filtering[C] // Proceedings of the 14th Conference on Uncertainty in Artificial Intelligence. San Francisco:Morgan Kaufmann, 1998: 43-52.
[6] CANDILLIER L, MEYER F, BOUllé M. Comparing state-of-the-art collaborative filtering systems[C] // Machine Learning and Data Mining in Pattern Recognition(MLDM’07). Berlin: Springerverlag Berlin, 2007: 548-562.
[7] PATRA B K, LAUNONEN R, OLLIKAINEN V, et al. A new similarity measure using Bhattacharyya coefficient for collaborative filtering in sparse data[J]. Knowledge-Based Systems, 2015, 82:163-177.
[8] KOUTRICA G, BERCOVITZ B, GARCIA H. FlexRecs: expressing and combining flexible recommendations[C] // Proceedings of the 35th SIGMOD International Conference on Management of Data Providence. New York: ACM, 2009: 745-757.
[9] CACHEDA F, CARNEIRO V, FERNÁNDEZ D, et al. Comparison of collaborative filtering algorithms: limitations of current techniques and proposals for scalable, high-performance recommender systems[J]. ACM Transactions on the Web(TWEB), 2011, 5(1):Article 2. Doi:10.1145/1921591.1921593.
[10] BOBADILLA J, SERRADILLA F, BERNAL J. A new collaborative filtering metric that improves the behavior of recommender systems[J]. Knowledge-Based Systems, 2010, 23:520-528.
[11] LIU HAIFENG, HU ZHENG, MIAN A, et al. A new user similarity model to improve the accuracy of collaborative filtering[J]. Knowledge-Based Systems, 2014, 56:156-166.
[12] SHARDANAND U. Social information filtering for music recommendation[D]. Boston: Massachusetts Institute of Technology, 1994.
[13] AHN H J. A new similarity measure for collaborative filtering to alleviate the new user cold-starting problem[J]. Information Sciences, 2008, 178(1):37-51.
[14] PIRASTEH P, HWANG D, JUNG J J. Exploiting matrix factorization to asymmetric user similarities in recommendation systems[J]. Knowledge-Based Systems, 2015, 83:51-57.
[15] JAMALI M, ESTER M. A matrix factorization technique with trust propagation for recommendation in social networks[C] // Proceedings of the 4th ACM conference on Recommender Systems(RecSys’10). New York: ACM, 2010: 135-142.
[16] SALAKHUTDINOV R, MNIH A. Probabilistic matrix factorization[C] // International Conference on Neural Information Processing Systems(NIPS’07). Cambridge: MIT, 2007: 1257-1264.
[17] 孙光福,吴乐,刘淇,等.基于时序行为的协同过滤推荐算法[J] ,软件学报,2013,24(11):2721-2733. SUN Guangfu, WU Le, LIU Qi, et al. Recommendations based on collaborative filtering by exploiting sequential behavior[J]. Journal of Software, 2013, 24(11):2721-2733.
[1] 张春英, 王立亚, 刘保相. 基于覆盖的区间概念格动态压缩原理与实现[J]. 山东大学学报(理学版), 2014, 49(08): 15-21.
[2] 倪展,吴群英*,施生塔. ND序列下最近邻密度估计的强相合速度[J]. J4, 2012, 47(12): 6-9.
[3] 蒋盛益1,庞观松2,张建军3. 基于聚类的垃圾邮件识别技术研究[J]. J4, 2011, 46(5): 71-76.
[4] 杨洋,王立宏*,刘其成. 一种主动式的半监督最近邻学习方法[J]. J4, 2011, 46(5): 110-115.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!