计算机科学 ›› 2019, Vol. 46 ›› Issue (8): 28-34.doi: 10.11896/j.issn.1002-137X.2019.08.005

• 大数据与数据科学* • 上一篇    下一篇

融合动态协同过滤和深度学习的推荐算法

邓存彬1,2, 虞慧群1, 范贵生1   

  1. (华东理工大学计算机科学与工程系 上海200237)1
    (上海市计算机软件测评重点实验室 上海201112)2
  • 收稿日期:2018-07-08 出版日期:2019-08-15 发布日期:2019-08-15
  • 通讯作者: 虞慧群(1967-),男,教授,博士生导师,CCF高级会员,主要研究方向为软件工程、形式化方法,E-mail:yhq@ecust.edu.cn
  • 作者简介:邓存彬(1993-),男,硕士生,主要研究方向为数据挖掘、机器学习;范贵生(1980-),男,副研究员,CCF会员,主要研究方向为软件工程、可信计算
  • 基金资助:
    国家自然科学基金(61702334,61772200),上海市浦江人才资助计划(17PJ1401900),上海市自然科学基金资助项目(17ZR1406900,17ZR1429700),华东理工大学教育科研基金(ZH1726108),上海应用技术学院资助合作创新基金会(XTCX2016-20)

Integrating Dynamic Collaborative Filtering and Deep Learning for Recommendation

DENG Cun-bin1,2, YU Hui-qun1, FAN Gui-sheng1   

  1. Department of Computer Science and Engineering,East China University of Science and Technology,Shanghai 200237,China)1
    (Shanghai Key Laboratory of Computer Software Evaluating and Testing,Shanghai 201112,China)2
  • Received:2018-07-08 Online:2019-08-15 Published:2019-08-15

摘要: 在信息爆炸的时代,推荐系统在减轻信息过载方面发挥了巨大的作用。目前,推荐系统普遍使用传统的协同过滤算法学习用户商品行为矩阵中的隐向量,但是其存在数据稀疏性和冷启动问题,同时未考虑用户的兴趣偏好以及商品的受欢迎程度会随时间发生改变,这极大地限制了推荐的准确性。已有学者利用深度学习模型学习辅助信息的特征来扩充协同过滤算法的特征,取得了一定的成果,但并未充分有效地解决全部问题。以电影推荐为研究对象,提出了融合动态协同过滤和深度学习的推荐算法。首先,利用动态协同过滤算法融入时间特征;然后,利用深度学习模型来学习用户和电影特征信息,以形成高维潜在空间的用户特征和电影特征的隐向量;最后,将其融入到动态协同过滤算法中。以MovieLens为实验数据集对电影的评分进行预测,实验结果表明所提算法提高了电影评分预测的准确性。

关键词: 电影推荐, 隐向量, 深度学习, 动态协同过滤

Abstract: In the era of information explosion,the recommendation system plays an enormous role in reducing information overload.At present,the recommendation system generally uses the traditional collaborative filtering algorithm to learn the hidden vector in the user-item behavior matrix,but it has the problem of data sparseness and cold start,and does not consider the customer preferences and the popularity dynamics of items.This greatly limits the accuracy of the recommendation system.Some scholars have used the deep learning model to learn the features of the auxiliary information to enrich the features of the collaborative filtering algorithm,and achieved certain results,which does not fully solve all the problems.This paper took film recommendation as the research object,and proposed a recommendation algorithm that combines dynamic collaborative filtering and deep learning.Firstly,the dynamic collaborative filtering algorithm incorporates temporal features.Secondly,it uses deep learning model to learn user and movie feature information to form the hidden vector of user features and movie features in high-dimensional latent space.Finally,it is integrated into the dynamic collaborative filtering algorithm.Extensive experiments on MovieLens datasets show that the proposed method improves the accuracy of film score prediction

Key words: Movie recommendation, Hidden vector, Deep learning, Dynamic collaborative filtering

中图分类号: 

  • TP311
[1] SUN H,HAN Z.An improved collaborative filtering algorithm for popular items of fusion items[J].Miniature Microcomputer Systems,2018,39(4):638-643.(in Chinese) 孙红,韩震.融合物品热门因子的协同过滤改进算法[J].小型微型计算机系统,2018,39(4):638-643.
[2] WENG X L,WANG Z J.Research progress of collaborative filtering recommendation algorithm[J].Computer Engineering and Applications,2018,54(1):25-31.(in Chinese) 翁小兰,王志坚.协同过滤推荐算法研究进展[J].计算机工程与应用,2018,54(1):25-31.
[3] XU R,ZHANG W.A recommendation system scoring prediction framework based on Adaboost algorithm[J].Journal of ComputerSystems,2017,26(8):107-113.(in Chinese) 徐日,张谧.基于Adaboost算法的推荐系统评分预测框架[J].计算机系统应用,2017,26(8):107-113.
[4] PORTEOUS I,ASUNCION A,WELLING M.Bayesian matrix factorization with side information and dirichlet process mixtures [C]∥Twenty-Fourth AAAI Conference on Artificial Intelligence.AAAI Press,2010:563-568.
[5] HUANG L W,JIANG B T,LU S Y,et al.A Survey of Recommendation Systems Based on Deep Learning [J].Chinese Journal of Computers,2018,41(7):191-219.(in Chinese) 黄立威,江碧涛,吕守业,等.基于深度学习的推荐系统研究综述[J].计算机学报,2018,41(7):191-219.
[6] ZHU Y,LI H,LIAO Y,et al.What to do next:modeling user behaviors by time-lstm [C]∥Twenty-Sixth International Joint Conference on Artificial Intelligence.2017:3602-3608.
[7] ZHENG L,NOROOZI V,YU P S.Joint deep modeling of users and items using reviews for recommendation[C]∥Proceedings of the Tenth ACM International Conference on Web Search and Data Mining.ACM,2017:425-434.
[8] CHENG H T,KOC L,HARMSEN J,et al.Wide & deep lear- ning for recommender systems[C]∥Proceedings of the 1st Workshop on Deep Learning for Recommender Systems.ACM,2016:7-10.
[9] QU Y,CAI H,REN K,et al.Product-based neural networks for user response prediction[C]∥2016 IEEE 16th International Conference on Data Mining (ICDM).IEEE,2016:1149-1154.
[10] HE X,LIAO L,ZHANG H,et al.Neural collaborative filtering[C]∥Proceedings of the 26th International Conference on World Wide Web.International World Wide Web Conferences Steering Committee,2017:173-182.
[11] ZHAO W,WANG W,YE J,et al.Leveraging long and short- term information in content-aware movie recommendation[J].arXiv:1712.09059,2017.
[12] KIM D,PARK C,OH J,et al.Convolutional matrix factorization for document context-aware recommendation[C]∥Proceedings of the 10th ACM Conference on Recommender Systems.ACM,2016:233-240.
[13] WEI J,HE J,CHEN K,et al.Collaborative filtering and deep learning based recommendation system for cold start items[J].Expert Systems with Applications,2017,69:29-39.
[14] WANG H,WANG N,YEUNG D Y.Collaborative deep learning for recommender systems[C]∥Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.ACM,2015:1235-1244.
[15] KOREN Y.Collaborative filtering with temporal dynamics[C]∥ Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining.ACM,2009:447-456.
[16] HARPER F M,KONSTAN J A.The movieLens datasets[J].Acm Transactions on Interactive Intelligent Systems,2016,5(4):1-19.
[17] KAWALE J,KAWALE J,FU Y.Deep collaborative filtering via marginalized denoising auto-encoder [C]∥ACM International on Conference on Information and Knowledge Management.ACM,2015:811-820.
[18] KIM D,PARK C,OH J,et al.Deep hybrid recommender systems via exploiting document context and statistics of items [J].Information Sciences,2017,417(C):72-87.
[1] 丁钰, 魏浩, 潘志松, 刘鑫. 网络表示学习算法综述[J]. 计算机科学, 2020, 47(9): 52-59.
[2] 何鑫, 许娟, 金莹莹. 行为关联网络:完整的变化行为建模[J]. 计算机科学, 2020, 47(9): 123-128.
[3] 叶亚男, 迟静, 于志平, 战玉丽, 张彩明. 基于改进CycleGan模型和区域分割的表情动画合成[J]. 计算机科学, 2020, 47(9): 142-149.
[4] 邓良, 许庚林, 李梦杰, 陈章进. 基于深度学习与多哈希相似度加权实现快速人脸识别[J]. 计算机科学, 2020, 47(9): 163-168.
[5] 暴雨轩, 芦天亮, 杜彦辉. 深度伪造视频检测技术综述[J]. 计算机科学, 2020, 47(9): 283-292.
[6] 袁野, 和晓歌, 朱定坤, 王富利, 谢浩然, 汪俊, 魏明强, 郭延文. 视觉图像显著性检测综述[J]. 计算机科学, 2020, 47(7): 84-91.
[7] 王文刀, 王润泽, 魏鑫磊, 漆云亮, 马义德. 基于堆叠式双向LSTM的心电图自动识别算法[J]. 计算机科学, 2020, 47(7): 118-124.
[8] 刘燕, 温静. 基于注意力机制的复杂场景文本检测[J]. 计算机科学, 2020, 47(7): 135-140.
[9] 张志扬, 张凤荔, 谭琪, 王瑞锦. 基于深度学习的信息级联预测方法综述[J]. 计算机科学, 2020, 47(7): 141-153.
[10] 蒋文斌, 符智, 彭晶, 祝简. 一种基于4Bit编码的深度学习梯度压缩算法[J]. 计算机科学, 2020, 47(7): 220-226.
[11] 陈晋音, 张敦杰, 林翔, 徐晓东, 朱子凌. 基于影响力最大化策略的抑制虚假消息传播的方法[J]. 计算机科学, 2020, 47(6A): 17-23.
[12] 程哲, 白茜, 张浩, 王世普, 梁宇. 使用深层卷积神经网络提高Hi-C 数据分辨率[J]. 计算机科学, 2020, 47(6A): 70-74.
[13] 赫磊, 邵展鹏, 张剑华, 周小龙. 基于深度学习的行为识别算法综述[J]. 计算机科学, 2020, 47(6A): 139-147.
[14] 孙正, 王新宇. 深度学习在光声成像中的应用现状[J]. 计算机科学, 2020, 47(6A): 148-152.
[15] 张曼, 李杰, 丁荣莉, 成昊天, 沈霁. 基于改进YOLO-V2算法的遥感图像目标检测技术研究[J]. 计算机科学, 2020, 47(6A): 176-180.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 雷丽晖,王静. 可能性测度下的LTL模型检测并行化研究[J]. 计算机科学, 2018, 45(4): 71 -75, 88 .
[2] 孙启,金燕,何琨,徐凌轩. 用于求解混合车辆路径问题的混合进化算法[J]. 计算机科学, 2018, 45(4): 76 -82 .
[3] 张佳男,肖鸣宇. 带权混合支配问题的近似算法研究[J]. 计算机科学, 2018, 45(4): 83 -88 .
[4] 伍建辉,黄中祥,李武,吴健辉,彭鑫,张生. 城市道路建设时序决策的鲁棒优化[J]. 计算机科学, 2018, 45(4): 89 -93 .
[5] 史雯隽,武继刚,罗裕春. 针对移动云计算任务迁移的快速高效调度算法[J]. 计算机科学, 2018, 45(4): 94 -99, 116 .
[6] 周燕萍,业巧林. 基于L1-范数距离的最小二乘对支持向量机[J]. 计算机科学, 2018, 45(4): 100 -105, 130 .
[7] 刘博艺,唐湘滟,程杰仁. 基于多生长时期模板匹配的玉米螟识别方法[J]. 计算机科学, 2018, 45(4): 106 -111, 142 .
[8] 耿海军,施新刚,王之梁,尹霞,尹少平. 基于有向无环图的互联网域内节能路由算法[J]. 计算机科学, 2018, 45(4): 112 -116 .
[9] 崔琼,李建华,王宏,南明莉. 基于节点修复的网络化指挥信息系统弹性分析模型[J]. 计算机科学, 2018, 45(4): 117 -121, 136 .
[10] 王振朝,侯欢欢,连蕊. 抑制CMT中乱序程度的路径优化方案[J]. 计算机科学, 2018, 45(4): 122 -125 .