| 402 | 4 | 36 |
| 下载次数 | 被引频次 | 阅读次数 |
用户对项目评分数据的稀疏性是影响推荐质量的主要因素之一,提出了融合评分数据和评论文本的深度学习模型,通过引入辅助信息缓解评分数据稀疏性的影响。利用评论文本可以获取用户的偏好信息和项目特征,而评分数据中又包含了用户和项目之间的潜在关联。现有的融合模型对评分数据的处理大多数都是采用矩阵分解方法,为了更好地利用评分数据中的有效信息,文中利用卷积神经网络处理评论文本,并引入注意力机制提取评论信息中具有代表性的评论,从而更好地表征用户偏好和项目特征。利用深度神经网络处理评分数据提取其中的深度特征,将特征进行融合来预测出用户对项目的评分。文中在Amazon数据集上进行验证,以均方误差MSE作为评价指标,结果表明所提出的模型优于多个优秀的基线模型。
Abstract：The sparsity of user rating data is one of the main factors affecting recommendation quality. A deep learning model combining rating data and comment text is proposed to mitigate the impact of rating data sparsity by introducing auxiliary information. The review text can be used to obtain the user's preference information and item characteristics, and the rating data contains the potential association between the user and the item. Most of the existing fusion models deal with rating data by using matrix factorization methods. In order to make better use of the effective information in the rating data, the convolutional neural network is used to process the comment text, and the attention mechanism is introduced to extract the representative comments from the comment information, so as to better characterize the user preferences and project characteristics. The deep neural network is used to process the score data to extract the deep features which are fused to predict the user's score on the item. It is verified on the Amazon data set with the mean square error(MSE) as the evaluation index. The results show that the proposed model is better than many excellent baseline models.
[1] 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):1619-1647.
[2] SUBRAMANIYASWAMY V,MANOGARAN G,LOGESH R,et al.An ontology-driven personalized food recommendation in IoT-based healthcare system[J].The Journal of Supercomputing,2018,75(6):3184-3216.
[3] HUANG Z H,SHAN G X,CHENG J J,et al.TRec:an efficient recommendation system for hunting passengers with deep neural networks[J].Neural Computing and Applications,2018,31:209-222.
[4] CHIEN T C,CHEN Z H,CHAN T W.Exploring long-term behavior patterns in a book recommendation system for reading[J].Journal of Educational Technology & Society,2017,20(2):27-36.
[5] GRBOVIC M,RADOSAVLJEVIC V,DJURIC N,et al.E-commerce in your inbox:product recommendations at scale[C]//Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining.Sydney,Australia:ACM,2015:1809-1818.
[6] 冷亚军，陆青，梁昌勇.协同过滤推荐技术综述[J].模式识别与人工智能，2014,27(8):720-734.
[7] 文俊浩，袁培雷，曾骏，等.基于标签主题的协同过滤推荐算法研究[J].计算机工程，2017,43(1):247-252.
[8] 单京晶.基于内容的个性化推荐系统研究[D].长春：东北师范大学，2015.
[9] KOREN Y,BELL R,VOLINSKY C.Matrix factorization techniques for recommender systems[J].Computer,2009,42(8):30-37.
[10] MNIH A,SALAKHUTDINOV R R.Probabilistic matrix factorization[C]//Proceedings of the advances in neural information processing systems.Vancouver,Canada:[s.n.],2008:1257-1264.
[11] LING G,LYU M R,KING I.Ratings meet reviews,a combined approach to recommend[C]//Proceedings of the 8th ACM conference on recommender systems.[s.l.]:ACM,2014:105-112.
[12] BLEI D M,NG A Y,JORDAN M I.Latent Dirichlet allocation[J].J.Mach.Learn.Res.,2003,3:993-1022.
[13] GANU G,KAKODKAR Y,MARIAN A.Improving the quality of predictions using textual information in online user reviews[J].Information Systems,2013,38(1):1-15.
[14] MCAULEY J,LESKOVEC J.Hidden factors and hidden topics:understanding rating dimensions with review text[C]//Proceedings of the 7th ACM conference on recommender systems.Hong Kong,China:ACM,2013:165-172.
[15] WANG S,MANNING C D.Baselines and bigrams:simple,good sentiment and topic classification[C]//Proceedings of the 50th annual meeting of the association for computational linguistic.Jeju Island,Korea:ACL,2012:90-94.
[16] COLLOBERT R,WESTON J.A unified architecture for natural language processing:deep neural networks with multitask learning[C]//Proceedings of the 25th international conference on machine learning (ICML '08).Helsinki,Finland:ACM,2008:160-167.
[17] 周飞燕，金林鹏，董军.卷积神经网络研究综述[J].计算机学报，2017,40(6):1229-1251.
[18] 高茂庭，徐彬源.基于循环神经网络的推荐算法[J].计算机工程，2019,45(8):198-202.
[19] 王永贵，尚庚.融合注意力机制的深度协同过滤推荐算法[J].计算机工程与应用，2019,55(13):8-14.
[20] AL-SABAHI K,ZUPING Z,NADHER M.A hierarchical structured self-attentive model for extractive document summarization (HSSAS)[J].IEEE Access,2018,6:24205-24212.
[21] SALAKHUTDINOV R,MNIH A,HINTON G.Restricted Boltzmann machines for collaborative filtering[C]//Proceedings of the 24th international conference on machine learning.Corvalis,USA:[s.n.],2007:791-798.
[22] WU Y,DUBOIS C,ZHENG A X,et al.Collaborative denoising auto-encoders for top-n recommender systems[C]//Proceedings of the ninth ACM international conference on web search and data mining.California,USA:ACM,2016:153-162.
[23] 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.Boston,USA:ACM,2016:233-240.
[24] 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.Cambridge,UK:ACM,2017:425-434.
[25] CHEN C,ZHANG M,LIU Y,et al.Neural attentional rating regression with review-level explanations[C]//Proceedings of the 2018 conference on world wide web.Lyon,France:[s.n.],2018:1583-1592.
基本信息:
中图分类号:TP391.3;TP18
引用信息:
[1]王艳,彭治,杜永萍.融合评分矩阵和评论文本的深度学习推荐模型[J].计算机技术与发展,2021,31(08):13-18.
基金信息:
国家重点研发计划(2019YFC1906002)
2021-08-10
2021-08-10