计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (6A): 220100066-8.doi: 10.11896/jsjkx.220100066

• 大数据&数据科学 • 上一篇    下一篇

一种时序情感记忆可约束可解释的序列推荐方法

郑麟1, 林艺璇1, 周东霖1, 朱福喜2   

  1. 1 汕头大学工学院 广东 汕头 515063;
    2 武汉大学计算机学院 武汉 430072
  • 出版日期:2023-06-10 发布日期:2023-06-12
  • 通讯作者: 郑麟(lzheng@stu.edu.cn)
  • 基金资助:
    国家自然科学基金(61902231);广东省基础与应用基础研究基金(2023A1515011240,2020A1515010531);广东教育科学规划课题高等教育专项(2021GXJK241)

Explainable Constraint Mechanism for Modeling Temporal Sentiment Memory in Sequential Recommendation

ZHENG Lin1, LIN Yixuan1, ZHOU Donglin1, ZHU Fuxi2   

  1. 1 College of Engineering,Shantou University,Shantou,Guangdong 515063,China;
    2 School of Computer Science,Wuhan University,Wuhan 430072,China
  • Online:2023-06-10 Published:2023-06-12
  • About author:ZHENG Lin,born in 1984,Ph.D,Lecturer,master supervisor,is a member of ACM and CCF.His main research interests include explainable deep learning and recommender systems.
  • Supported by:
    National Natural Science Foundation of China(61902231),Guangdong Basic and Applied Basic Research Foundation(2023A1515011240,2020A1515010531) and Higher Education Special Project of Guangdong Education Science Planning(2021GXJK241).

PDF (PC)

332

摘要： 序列推荐研究近年来在推荐领域中发展迅速,已有的序列推荐方法善于捕捉用户的时序行为来实现偏好预测。其中,一些先进的方法融入用户的情感信息来引导行为挖掘。然而,先进的基于情感的序列推荐模型未考虑对多类别的用户情感序列进行关联挖掘;并且,这类方法无法直观地解释时序情感对用户偏好的贡献。为了弥补上述方法的局限,本工作首次尝试以记忆体的形式存储时序情感并对其施加约束。具体地,文中提出了情感自我约束和情感相互约束两种机制,来挖掘多类别情感之间的关联并辅助用户行为完成序列推荐。进一步地,提出的记忆框架能记录用户的时序情感注意力,从而在准确预测用户时序偏好的基础上提供一定程度的直观解释。实验结果表明,所提方法的性能优于先进的序列推荐方法,并且比基于情感的序列推荐模型具有更好的可解释效果。

关键词: 序列推荐, 时序情感, 情感记忆, 情感约束, 可解释推荐

Abstract: In recent years,the research of sequential recommendation has developed rapidly in the recommendation field,existing methods are good at capturing users’ sequential behavior to achieve preference prediction.Among them,some advanced methods integrate users’ sentiment information to guide behavior mining.However,the advanced sentiment-based models do not consider mining relations between multi-category user sentiment sequences.Moreover,such methods cannot intuitively explain the contribution of temporal sentiments to user preferences.To make up for the above shortcomings,this paper first attempts to store temporal sentiments in the form of memory and impose constraints on them.Specifically,this research proposes two mechanisms including sentiment self-constraint and sentiment mutual-constraint to explore the associations between multiple categories of sentiments and assist user behaviors in completing sequential recommendations.Furthermore,the proposed memory framework is able to record users’ temporal sentiment attention,so that it can provide a certain degree of intuitive explanation on the basis of accurately predicting users’ temporal preference.Experimental results show that our approach outperforms existing state-of-the-art sequential methods,and it has better explainable effects than the sentiment-based sequential recommendation models.

Key words: Sequential recommendation, Temporal sentiment, Sentiment memory, Sentiment constraint, Explainable recommendation

中图分类号: 

  • TP311
[1]ZHENG L,ZHU F X,YAO X.Recommendation rating prediction based on attribute boosting with partial sampling[J].Chinese Journal of Computers,2016,39(8):1501-1514.
[2]WANG S,HU L,WANG Y,et al.Sequential RecommenderSystems:Challenges,Progress and Prospects[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence(IJCAI’19).2019:6332-6338.
[3]HIDASI B,KARATZOGLOU A,BALTRUNAS L,et al.Session-based Recommendations with Recurrent Neural Networks[C]//Proceedings of the 4th International Conference on Lear-ning Representations(ICLR ’16).2016:1-10.
[4]TAN Y K,XU X,LIU Y.Improved Recurrent Neural Networks for Session-based Recommendations[C]//Proceedings of the 1st Workshop on Deep Learning for Recommender Systems(DLRS’16).2016:17-22.
[5]KANG W,MCAULEY J.Self-Attentive Sequential Recommendation[C]//Proceedings of the 2018 IEEE International Confe-rence on Data Mining(ICDM’18).2018:197-206.
[6]HUANG J,ZHAO W X,DOU H,et al.Improving sequential recommendation with knowledge-enhanced memory networks[C]//Proceedings of the 41st International ACM SIGIR Confe-rence on Research & Development in Information Retrieval(SIGIR’18).2018:505-514.
[7]PAPPAS N,POPESCU-BELIS A.Adaptive sentiment-aware one-class collaborative filtering[J].Expert Systems with Applications,2016,43(1):23-41.
[8]XU Y,YANG Y,HAN J,et al.NeuO:Exploiting the sentimental bias between ratings and reviews with neural networks.Neural Networks,2019,111(3):77-88.
[9]LI G,CHEN Q,ZHENG B,et al.Time-aspect-sentiment Re-commendation Models Based on Novel Similarity Measure Me-thods[J].ACM Transactions on the Web,2020,14(2):1-26.
[10]ZHENG L,GUO N,CHEN W,et al.Sentiment-Guided Sequential Recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research & Development in Information Retrieval(SIGIR’20).2020:1957-1960.
[11]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isAll you Need[C]//Proceedingsof the 31st International Confe-rence on Neural Information Processing Systems(NIPS’17).2017:6000-6010.
[12]CHILD R,GRAY S,RADFORD A,et al.Generating Long Sequences with Sparse Transformers[J].arXiv:1904.10509,2019.
[13]RENDLE S,FREUDENTHALER C,SCHMIDT-THIEME L.Factorizing personalized Markov chains for next-basket recommendation[C]//Proceedings of the 19th International Confe-rence on World Wide Web(WWW’10).2010:811-820.
[14]WANG P,GUO J,LAN Y,et al.Learning hierarchical representation model for next basket recommendation[C]//Proceedings of the 38th International ACM SIGIR Conference on Research &Development in Information Retrieval(SIGIR’15).2015:403-412.
[15]SUKHBAATAR S,SZLAM A,WESTON J,et al.End-to-end memory networks[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems-Volume 2(NIPS’15).2015:2440-2448.
[16]EBESU T,SHEN B,FANG Y.Collaborative memory network for recommendation systems[C]//Proceedings of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval(SIGIR’18).2018:515-524.
[17]CHEN X,XU H,ZHANG Y,et al.Sequential recommendation with user memory networks[C]//Proceedings of the 11th ACM International Conference on Web Search and Data Mining(WSDM’18).2018:108-116.
[18]WANG Y G,JIANG C C,FENG X N,et al.Time Aware Point-of-interest Recommendation[J].Computer Science,2021,48(9):43-49.
[19]ZHAO J L,ZHAO Z Y.Recommendation Algorithm Based on Heterogeneous Information Network Embedding and Attention Neural Network[J].Computer Science,2021,48(8):72-79.
[20]ZHANG T,ZHAO P,LIU Y,et al.Feature-level Deeper Self-Attention Network for Sequential Recommendation[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence(IJCAI’19).2019:4320-4326.
[21]CHEN Q,CHE M M,GUO X,et al.Recurrent Convolution Attention Model for Sentiment Classification[J].Computer Science,2021,48(2):245-249.
[22]SUN Z Q,LUO Y L,ZHENG X Y,et al.Intelligent TravelRoute Recommendation Method Integrating User Emotion and Similarity[J].Computer Science,2021,48(S1):226-230.
[23]CHRISTOPHER M,MIHAI S,JOHN B,et al.The StanfordCoreNLP Natural Language Processing Toolkit[C]//Procee-dings of the 52nd Annual Meeting of the Association for Computational Linguistics:System Demonstrations(ACL’14).2014:55-60.
[24]MENG X,WANG S,LIU H Z,et al.Exploiting Emotion on Reviews for Recommender Systems[C]//Proceedingsof the 32nd AAAI Conference on Artificial Intelligence(AAAI’18).2018:3788-3795.
[25]HYUN D,PARK C,YANG M C,et al.Review Sentiment-Guided Scalable Deep Recommender System[C]//Proceedings of the 41st International ACM SIGIR Conference on Research &Development in Information Retrieval(SIGIR’18).2018:965-968.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发