计算机科学 ›› 2019, Vol. 46 ›› Issue (8): 111-115.doi: 10.11896/j.issn.1002-137X.2019.08.018

• 2018 全国高性能计算学术年会 • 上一篇    下一篇

基于用户向量化表示和注意力机制的深度神经网络推荐模型

郭旭1, 朱敬华1,2   

  1. (黑龙江大学计算机科学技术学院 哈尔滨150080)1
    (黑龙江省数据库与并行计算重点实验室 哈尔滨150080)2
  • 收稿日期:2018-10-21 发布日期:2019-08-15
  • 通讯作者: 朱敬华(1976-),男,教授,CCF会员,主要研究方向为数据挖掘、推荐系统,E-mail:zhujinghua@hlju.edu.cn
  • 作者简介:郭旭(1989-),男,硕士,主要研究方向为数据挖掘
  • 基金资助:
    黑龙江省教育厅科研项目(12531498)

Deep Neural Network Recommendation Model Based on User Vectorization Representation and Attention Mechanism

GUO Xu1, ZHU Jing-hua1,2   

  1. (School of Computer Science and Technology,Heilongjiang University,Harbin 150080,China)1
    (Key Laboratory of Database and Parallel Computing of Heilongjiang Province,Harbin 150080,China)2
  • Received:2018-10-21 Published:2019-08-15

摘要: 随着互联网应用的蓬勃发展,推荐系统作为解决信息过载的有效手段,成为了工业界与学术界的研究热点。面向用户隐式反馈的传统推荐算法主要基于协同过滤和排序学习等方法,但这些方法未充分利用用户行为中的隐式反馈特征。文中提出了一种基于神经网络的用户向量化表示模型,其能够充分利用用户的异构的隐式反馈行为特征。同时,借鉴机器翻译中的self-attention机制,设计了一种神经注意力推荐模型,其融合用户向量化表示和用户-项目交互的动态时序特征以提高推荐系统的性能。在公开数据集上进行对比实验,通过召回率、准确率、NDCG 3个指标评价推荐性能。结果表明,与其他面向隐式反馈的推荐模型相比,所提推荐模型具有更好的推荐性能,并且对用户行为特征具有很好的泛化能力。

关键词: 推荐系统, 神经网络, 注意力机制, 隐式反馈, 深度学习

Abstract: With the rapid development of Internet application,recommendation system,as an effective measure to solve information overloading,has become a research hot topic in industry and academia.Traditional recommendation algorithms for users’ implicit feedback are mainly based on collaborative filtering and learning-to-rank method,which do not make full use of the implicit feedback features in users’ behaviors.In this paper,a users’ vectorization representation model based on neural network was proposed,which can make full use of heterogeneous implicit feedback features of users’ behaviors.At the same time,referring to the self-attention mechanism in machine translation,this paper designed a neural attentive recommendation model which integrates the dynamic temporal features of user-item interaction and user vectorization representation,to improve the performance of the recommendation system.The comparison experiment is conducted on two public datasets,and the recommendation performance is evaluated by recall,precision and NDCG.Compared with other recommendation models for implicit feedback,the proposed recommendation model has better recommendation performance and better generalization ability

Key words: Recommendation system, Neural networks, Attention mechanism, Implicit feedback, Deep learning

中图分类号: 

  • TP391
[1] VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[M]∥Advances in Neural Information Processing Systems.Bertin:Springer,2017:5998-6008.
[2] SARWAR B,KARYPIS G,KONSTAN J,et al.Item-based collaborative filtering recommendation algorithms[C]∥Procee-dings of the 10th International Conference on World Wide Web.ACM,2001:285-295.
[3] OSTUNI V C,DI NOIA T,DI SCIASCIO E,et al.Top-n recommendations from implicit feedback leveraging linked open data[C]∥Proceedings of the 7th ACM Conference on Recommender Systems.ACM,2013:85-92.
[4] ZIMDARS A,CHICKERING D M,MEEK C.Using temporal data for making recommendations[C]∥Proceedings of the Se-venteenth Conference on Uncertainty in artificial intelligence.Morgan Kaufmann Publishers Inc.,2001:580-588.
[5] RENDLE S,FREUDENTHALER C,GANTNER Z,et al.BPR:Bayesian personalized ranking from implicit feedback[C]∥Proceedings of the Twenty-fifth Conference on Uncertainty in Artificial Intelligence.AUAI Press,2009:452-461.
[6] SEDHAIN S,MENON A K,SANNER S,et al.Autorec:Autoencoders meet collaborative filtering[C]∥Proceedings of the 24th International Conference on World Wide Web.ACM,2015:111-112.
[7] ELKAHKY A M,SONG Y,HE X.A multi-view deep learning approach for cross domain user modeling in recommendation systems[C]∥Proceedings of the 24th International Conference on World Wide Web.International World Wide Web Confe-rences Steering Committee,2015:278-288.
[8] 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.
[9] HIDASI B,KARATZOGLOU A,BALTRUNAS L,et al.Session-based recommendations with recurrent neural networks[J].arXiv:1511.06939,2015.
[10] GONG Y,ZHANG Q.Hashtag Recommendation Using Attention-Based Convolutional Neural Network[C]∥Proceedings of International Joint Conference on Artificial Intelligence.AAAI Press,2016:2782-2788.
[11] LI Y,LIU T,JIANG J,et al.Hashtag recommendation with topical attention-based LSTM[C]∥Proceedings of the 26th International Conference on Computational Linguistics.ACL,2016:3019-3029.
[12] XIAO J,YE H,HE X,et al.Attentional factorization machines:Learning the weight of feature interactions via attention networks[J].arXiv:1708.04617,2017.
[1] 周燕, 曾凡智, 吴臣, 罗粤, 刘紫琴. 基于深度学习的三维形状特征提取方法[J]. 计算机科学, 2019, 46(9): 47-58.
[2] 马露, 裴伟, 朱永英, 王春立, 王鹏乾. 基于深度学习的跌倒行为识别[J]. 计算机科学, 2019, 46(9): 106-112.
[3] 李青华, 李翠平, 张静, 陈红, 王绍卿. 深度神经网络压缩综述[J]. 计算机科学, 2019, 46(9): 1-14.
[4] 王鑫, 孟浩浩, 姜小涛, 陈胜勇, 孙凌云. 基于神经网络的角色运动合成研究进展[J]. 计算机科学, 2019, 46(9): 22-27.
[5] 王嫣然, 陈清亮, 吴俊君. 面向复杂环境的图像语义分割方法综述[J]. 计算机科学, 2019, 46(9): 36-46.
[6] 孙中锋, 王静. 用于基于方面情感分析的RCNN-BGRU-HN网络模型[J]. 计算机科学, 2019, 46(9): 223-228.
[7] 石春丹, 秦岭. 基于BGRU-CRF的中文命名实体识别方法[J]. 计算机科学, 2019, 46(9): 237-242.
[8] 缪永伟, 李高怡, 鲍陈, 张旭东, 彭思龙. 基于卷积神经网络的图像局部风格迁移[J]. 计算机科学, 2019, 46(9): 259-264.
[9] 江泽涛, 秦嘉奇, 胡硕. 基于多路卷积神经网络的多光谱场景识别方法[J]. 计算机科学, 2019, 46(9): 265-270.
[10] 石晓红, 黄钦开, 苗佳欣, 苏卓. 基于卷积网络的边缘保持滤波方法[J]. 计算机科学, 2019, 46(9): 277-283.
[11] 邓存彬, 虞慧群, 范贵生. 融合动态协同过滤和深度学习的推荐算法[J]. 计算机科学, 2019, 46(8): 28-34.
[12] 杜威, 丁世飞. 多智能体强化学习综述[J]. 计算机科学, 2019, 46(8): 1-8.
[13] 张艳红, 张春光, 周湘贞, 王怡鸥. 项目多属性模糊联合的多样性视频推荐算法[J]. 计算机科学, 2019, 46(8): 78-83.
[14] 张义杰, 李培峰, 朱巧明. 基于自注意力机制的事件时序关系分类方法[J]. 计算机科学, 2019, 46(8): 244-248.
[15] 方杰, 李培峰, 朱巧明. 基于多注意力机制的事件同指消解方法[J]. 计算机科学, 2019, 46(8): 277-281.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 夏庆勋,庄毅. 一种基于局部性原理的远程验证机制[J]. 计算机科学, 2018, 45(4): 148 -151, 162 .
[2] 刘琴. 计算机取证过程中基于约束的数据质量问题研究[J]. 计算机科学, 2018, 45(4): 169 -172 .
[3] 钟菲,杨斌. 基于主成分分析网络的车牌检测方法[J]. 计算机科学, 2018, 45(3): 268 -273 .
[4] 王帅,刘娟,毕姚姚,陈哲,郑群花,段慧芳. 基于两步聚类和随机森林的乳腺腺管自动识别方法[J]. 计算机科学, 2018, 45(3): 247 -252 .
[5] 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 .
[6] 陆佳炜,马俊,张元鸣,肖刚. 面向全局社交服务网的Web服务聚类方法[J]. 计算机科学, 2018, 45(3): 204 -212 .
[7] 冉正,罗蕾,晏华,李允. AUTOSAR可运行实体-任务自动映射方法研究[J]. 计算机科学, 2018, 45(4): 190 -195, 226 .
[8] 丁舒阳,黎冰,侍洪波. 基于改进的离散PSO算法的FJSP的研究[J]. 计算机科学, 2018, 45(4): 233 -239, 256 .
[9] 侯彦娥,孔云峰,党兰学. 求解多车型校车路径问题的混合集合划分的GRASP算法[J]. 计算机科学, 2018, 45(4): 240 -246 .
[10] 李昊阳,符云清. 基于标签聚类与项目主题的协同过滤推荐算法[J]. 计算机科学, 2018, 45(4): 247 -251 .