Please wait a minute...
Advanced Search
数据分析与知识发现  2019, Vol. 3 Issue (11): 43-51     https://doi.org/10.11925/infotech.2096-3467.2019.0370
     研究论文 本期目录 | 过刊浏览 | 高级检索 |
基于用户多类型兴趣波动趋势预测分析的个性化推荐方法 *
丁浩,李树青()
南京财经大学信息工程学院 南京 210023
Personalized Recommendation Based on Predictive Analysis of User’s Interests
Hao Ding,Shuqing Li()
School of Information Engineering, Nanjing University of Finance and Economics, Nanjing 210023, China
全文: PDF (935 KB)   HTML ( 14
输出: BibTeX | EndNote (RIS)      
摘要 

【目的】构建一种基于用户多类型兴趣波动特征预测的推荐方法以提升推荐效果。【方法】针对每种类型用户评分数据加入时间衰减因子并使用神经网络对数据波动线性拟合, 选择最优参数结果并对比评估方法有效性。【结果】通过5组不同的用户数据进行仿真实验, 结果表明, 本文方法预测结果的MAE和RMSE分别较对比方法最高降低幅度达到47.63%和44.61%。【局限】由于时间波动的分析依赖用户历史数据, 当历史数据量过于稀疏时需采用额外冷启动算法对数据进行预处理。【结论】该方法结合用户对不同商品类型兴趣漂移特征的波动分析和预测, 使推荐结果更准确。

服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
丁浩
李树青
关键词 时间序列兴趣类型波动分析个性化推荐    
Abstract

[Objective] This paper tries to construct a time series prediction model based on the fluctuation of users’ historical interests, aiming to improve the recommendation results. [Methods] We added time attenuation factor to the ratings by each type of users and linearly fit the data fluctuation with neural network. Then, we chose the optimal parameters to compare the effectiveness of the proposed method. [Results] We conducted five rounds of user simulation tests and found the MAE and RMSE errors of the proposed method were reduced by 47.63% and 44.61%. [Limitations] Analysis of time fluctuation relies on users’ historical data, thus, additional cold-start algorithm is needed to preprocess the data. [Conclusions] The proposed method could effectively analyze and predict the changing of interests in different commodities, and provide more accurate recommendation lists.

Key wordsTime Series    Interest Type    Fluctuation Analysis    Personalized Recommendation
收稿日期: 2019-04-08      出版日期: 2019-12-18
ZTFLH:  TP391  
基金资助:*本文系江苏省研究生科研与实践创新计划项目“大数据场景下用户兴趣模式演变趋势分析”(项目编号: KYCX18_1391);国家社会科学基金项目“基于大数据分析的数字图书馆个性化服务模式创新研究”(项目编号: 16BTQ030)
通讯作者: 李树青     E-mail: leeshuqing@163.com
引用本文:   
丁浩,李树青. 基于用户多类型兴趣波动趋势预测分析的个性化推荐方法 *[J]. 数据分析与知识发现, 2019, 3(11): 43-51.
Hao Ding,Shuqing Li. Personalized Recommendation Based on Predictive Analysis of User’s Interests. Data Analysis and Knowledge Discovery, 2019, 3(11): 43-51.
链接本文:  
https://manu44.magtech.com.cn/Jwk_infotech_wk3/CN/10.11925/infotech.2096-3467.2019.0370      或      https://manu44.magtech.com.cn/Jwk_infotech_wk3/CN/Y2019/V3/I11/43
  模型框架示意图
用户数量 电影数量 评分数量 稀疏度(%)
138 493 27 278 20 000 263 0.5294
  数据集规模
Genre\Set Set#1 Set#2 Set#3 Set#4 Set#5 Total
Train Test Train Test Train Test Train Test Train Test Train Test
Action 31 344 11 022 30 990 11 147 27 434 9 686 30 954 10 891 30 661 10 924 151 383 53 670
Adventure 24 134 8 561 24 306 8 679 21 295 7 528 23 980 8 611 24 704 8 763 118 419 42 142
Animation 5 576 2 559 5 661 2 533 5 206 2 275 5 707 2 520 6 046 2 722 28 196 12 609
Children 9 317 2 945 9 180 2 988 8 220 2 602 8 907 2 819 10 115 3 122 45 739 14 476
Comedy 42 591 14 162 42 343 13 892 37 222 12 000 40 770 13 269 44 719 14 557 207 645 67 880
Crime 18 343 6 071 18 627 6 178 16 251 5 478 17 945 5 953 18 248 6 207 89 414 29 887
Documentary 906 800 897 799 760 750 875 786 941 848 4 379 3 983
Drama 48 023 16 793 49 602 17 172 42 774 15 218 47 443 16 580 49 783 17 723 237 625 83 486
Fantasy 11 197 4 472 11 210 4 498 9 873 3 881 10 979 4 357 11 578 4 562 54 837 21 770
Film-Noir 1 274 276 1 223 324 1 063 276 1 278 306 1 324 316 6 162 1 498
Horror 9 156 3 095 7 674 2 619 7 544 2 552 7 678 2 655 8 451 2 707 40 503 13 628
IMAX 2 175 1 424 2 276 1 485 1 825 1 268 2 130 1 374 2 391 1 491 10 797 7 042
Musical 4 828 1 560 5 025 1 609 4 298 1 443 4 814 1 440 5 256 1 710 24 221 7 762
Mystery 8 895 2 961 8 707 2 929 7 734 2 624 8 494 2 827 8 817 2 966 42 647 14 307
Romance 21 445 6 704 22 305 7 008 18 970 5 936 21 486 6 452 23 009 7 135 107 215 33 235
Sci-Fi 18 223 6 129 17 597 5 988 15 823 5 342 17 444 6 001 17 545 5 949 86 632 29 409
Thriller 30 174 10 228 29 597 10 126 26 384 9 126 28 868 9 893 29 582 10 154 144 605 49 527
War 5 959 1 732 6 025 1 836 5 119 1 520 6 049 1 816 6 251 1 855 29 403 8 759
Western 2 608 893 2 357 805 2 085 693 2 360 798 2 399 870 11 809 4 059
Total 296 168 102 387 295 602 102 615 259 880 90 198 288 161 99 348 301 820 104 581 1 441 631 499 129
  测试数据划分标签数量
  每种迭代参数θ对预测的影响
  不同分组的测试结果比较
评估方法 运行时间(s) 响应时间(s)
NormalPredictor 0.15 0.18
KNNBasic 0.50 2.79
SVD++ 558.98 9.81
NMF 7.04 0.19
TGE-CF 1.15 0.15
  每次迭代的运行时间和响应时间
[1] Goldberg D, Nichols D, Oki B M , et al. Using Collaborative Filtering to Weave an Information Tapestry[J]. Communications of the ACM, 1992,35(12):61-71.
[2] Ma H, King I, Lyu M R . Learning to Recommend with Social Trust Ensemble [C]// Proceedings of the 32nd International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM, 2009: 203-210.
[3] Konstas I, Stathopoulos V, Jose J M . On Social Networks and Collaborative Recommendation [C]// Proceedings of the 32nd International ACM SIGIR Conference on Research & Development in Information Retrieval. ACM, 2009: 195-202.
[4] Liu Z, Tan K, Wang X Q , et al. A Learning Framework for Temporal Recommendation Without Explicit Iterative Optimization[J]. Applied Soft Computing, 2018,67:529-539.
doi: 10.1016/j.asoc.2018.03.024
[5] Rendle S, Freudenthaler C, Schmidt-Thieme L . Factorizing Personalized Markov Chains for Next-Basket Recommendation [C]// Proceedings of the 19th International Conference on World Wide Web. Springer, 2010: 811-820.
[6] 印鉴, 王智圣, 李琪 , 等. 基于大规模隐式反馈的个性化推荐[J]. 软件学报, 2014,25(9):1953-1966.
[6] ( Yin Jian, Wang Zhisheng, Li Qi , et al. Personalized Recommendation Based on Large-Scale Implicit Feedback[J]. Journal of Software, 2014,25(9):1953-1966.)
[7] Jamali M, Ester M . A Matrix Factorization Technique with Trust Propagation for Recommendation in Social Networks [C]// Proceedings of the 2010 ACM Conference on Recommender Systems. ACM, 2010: 135-142.
[8] Liu J, Wu C, Xiong Y , et al. List-Wise Probabilistic Matrix Factorization for Recommendation[J]. Information Sciences, 2014,278:434-447.
doi: 10.1016/j.ins.2014.03.063
[9] Koren Y . Collaborative Filtering with Temporal Dynamics[J]. Communications of the ACM, 2010,53(4):89-97.
doi: 10.3389/fphys.2019.00182 pmid: 30971864
[10] Hosseini S, Alizadeh K, Khodadadi A , et al. Recurrent Poisson Factorization for Temporal Recommendation [C]// Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 2017: 847-855.
[11] Xu Z, Yang Y, Hauptmann A G . A Discriminative CNN Video Representation for Event Detection [C]// Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2014: 1798-1807.
[12] Graves A, Mohamed A R, Hinton G . Speech Recognition with Deep Recurrent Neural Networks [C]// Proceedings of the 2013 IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2013: 6645-6649.
[13] Weng C, Yu D, Watanabe S , et al. Recurrent Deep Neural Networks for Robust Speech Recognition [C]// Proceedings of the 2014 IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2014: 5532-5536.
[14] Arevian G, Panchev C. Optimising the Hystereses of a Two Context Layer RNN for Text Classification[C]// Proceedings of the 2007 IEEE International Joint Conference on Neural Networks. IEEE Press, 2007: 2936-2941.
[15] Wang C, Jiang F, Yang H . A Hybrid Framework for Text Modeling with Convolutional RNN [C]// Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. ACM, 2017: 2061-2070.
[16] Selvin S, Vinayakumar R, Gopalakrishnan E A , et al. Stock Price Prediction Using LSTM, RNN and CNN-Sliding Window Model [C]// Proceedings of the 2017 International Conference on Advances in Computing. IEEE, 2017: 1643-1647.
[17] Madan R, Sarathimangipudi P . Predicting Computer Network Traffic: A Time Series Forecasting Approach Using DWT, ARIMA and RNN [C]//Proceedings of the 11th International Conference on Contemporary Computing. IEEE, 2018: 1-5.
[18] Yu F, Liu Q, Wu S , et al. A Dynamic Recurrent Model for Next Basket Recommendation [C]// Proceedings of the 39th International ACM SIGIR Conference. ACM, 2016: 729-732.
[19] 田垅, 刘宗田 . 最小二乘法分段直线拟合[J]. 计算机科学, 2012,39(S1):482-484.
[19] ( Tian Long, Liu Zongtian . Least- squares Method Piecewise Linear Fitting[J]. Computer Science, 2012,39(S1):482-484.)
[20] Bottou L . Large-Scale Machine Learning with Stochastic Gradient Descent [C]//Proceedings of COMPSTAT’2010. 2010: 177-186.
[21] Chai T, Draxler R R . Root Mean Square Error (RMSE) or Mean Absolute Error (MAE)? - Arguments Against Avoiding RMSE in the Literature[J]. Geoscientific Model Development, 2014,7(3):1247-1250.
doi: 10.5194/gmd-7-1247-2014
[1] 吴彦文, 蔡秋亭, 刘智, 邓云泽. 融合多源数据和场景相似度计算的数字资源推荐研究*[J]. 数据分析与知识发现, 2021, 5(11): 114-123.
[2] 丁浩, 艾文华, 胡广伟, 李树青, 索炜. 融合用户兴趣波动时序的个性化推荐模型*[J]. 数据分析与知识发现, 2021, 5(11): 45-58.
[3] 魏伟,郭崇慧,邢小宇. 基于语义关联规则的试题知识点标注及试题推荐*[J]. 数据分析与知识发现, 2020, 4(2/3): 182-191.
[4] 颜靖华,侯苗苗. 基于LSTM网络的盗窃犯罪时间序列预测研究*[J]. 数据分析与知识发现, 2020, 4(11): 84-91.
[5] 张怡文,张臣坤,杨安桔,计成睿,岳丽华. 基于条件型游走的四部图推荐方法*[J]. 数据分析与知识发现, 2019, 3(4): 117-125.
[6] 叶佳鑫,熊回香. 基于标签的跨领域资源个性化推荐研究*[J]. 数据分析与知识发现, 2019, 3(2): 21-32.
[7] 聂卉. 结合词向量和词图算法的用户兴趣建模研究 *[J]. 数据分析与知识发现, 2019, 3(12): 30-40.
[8] 李杰, 杨芳, 徐晨曦. 考虑时间动态性和序列模式的个性化推荐算法*[J]. 数据分析与知识发现, 2018, 2(7): 72-80.
[9] 侯银秀, 李伟卿, 王伟军, 张婷婷. 基于用户偏好与商品属性情感匹配的图书个性化推荐研究*[J]. 数据分析与知识发现, 2017, 1(8): 9-17.
[10] 陈梅梅, 薛康杰. 基于标签簇多构面信任关系的个性化推荐算法研究*[J]. 数据分析与知识发现, 2017, 1(5): 94-101.
[11] 陈梅梅, 薛康杰. 基于改进张量分解模型的个性化推荐算法研究*[J]. 数据分析与知识发现, 2017, 1(3): 38-45.
[12] 谭学清,张磊,黄翠翠,罗琳. 融合领域专家信任与相似度的协同过滤推荐算法研究*[J]. 现代图书情报技术, 2016, 32(7-8): 101-109.
[13] 谢琪,崔梦天. 基于相似性群体的混合型Web服务推荐*[J]. 现代图书情报技术, 2016, 32(6): 80-87.
[14] 林园园,战洪飞,余军合,李长江,张凡. 基于产品评论的消费者情感波动分析模型构建及实证研究*[J]. 现代图书情报技术, 2016, 32(11): 44-53.
[15] 祝婷, 秦春秀, 李祖海. 基于用户分类的协同过滤个性化推荐方法研究[J]. 现代图书情报技术, 2015, 31(6): 13-19.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
版权所有 © 2015 《数据分析与知识发现》编辑部
地址:北京市海淀区中关村北四环西路33号 邮编:100190
电话/传真:(010)82626611-6626,82624938
E-mail:jishu@mail.las.ac.cn