融合评论文本特征和评分图卷积表示的推荐模型

doi:10.11959/j.issn.1000-436x.2022049

通信学报 ›› 2022, Vol. 43 ›› Issue (3): 164-171.doi: 10.11959/j.issn.1000-436x.2022049

融合评论文本特征和评分图卷积表示的推荐模型

冯海林, 张潇, 刘同存

浙江农林大学数学与计算机科学学院，浙江杭州 311300

修回日期:2022-02-14 出版日期:2022-03-25 发布日期:2022-03-01
作者简介:冯海林（1980- ），男，安徽安庆人，博士，浙江农林大学教授、博士生导师，主要研究方向为计算机视觉、智能信息处理和物联网
张潇（1997- ），女，浙江台州人，浙江农林大学硕士生，主要研究方向为深度学习和推荐算法
刘同存（1983- ），男，山东临沂人，博士，浙江农林大学讲师，主要研究方向为数据挖掘、视频理解、深度学习和推荐算法
基金资助:
国家自然科学基金资助项目(U1809208);浙江省自然科学基金资助项目(LGG22F020010)

Recommendation model combining review’s feature and rating graph convolutional representation

Hailin FENG, Xiao ZHANG, Tongcun LIU

College of Mathematics and Computer Science, Zhejiang A ＆F University, Hangzhou 311300, China

Revised:2022-02-14 Online:2022-03-25 Published:2022-03-01
Supported by:
The National Natural Science Foundation of China(U1809208);The Natural Science Foundation of Zhejiang Province(LGG22F020010)

摘要/Abstract

摘要：

为了充分利用评分的有效信息，并进一步研究评论的重要性，提出了一种融合评论文本特征和评分图卷积表示的推荐模型，利用图卷积编码学习用户和商品在评分上的特征表示，结合文本卷积特征，使用注意力机制来区分评论的重要性，然后通过隐因子模型把在评论和评分上学习到的特征表示融合产生推荐。在亚马逊公开数据集上的实验结果表明，提出的模型显著优于现有的模型，证明了提出的模型的有效性。

关键词: 推荐模型, 图卷积编码, 注意力机制, 隐因子模型

Abstract:

In order to fully exploit the effective information of the ratings and further investigate the importance of the review, a recommendation model combining review’s feature and rating graph convolutional representation was proposed.Graph convolutional neural network was used to learn the representation of user and item from the ratings data.Combining with text convolutional features, attention mechanism was utilized to distinguish the importance of the review.Finally, the representation learned from the review and the rating data was fused by the hidden factor model.The experimental results on Amazon’s public data showed that the proposed model significantly outperformed the traditional approaches, proving the effectiveness of the proposed model.

Key words: recommender model, graph convolutional encoder, attention mechanism, latent factor model

中图分类号:

TP391

冯海林, 张潇, 刘同存. 融合评论文本特征和评分图卷积表示的推荐模型[J]. 通信学报, 2022, 43(3): 164-171.

Hailin FENG, Xiao ZHANG, Tongcun LIU. Recommendation model combining review’s feature and rating graph convolutional representation[J]. Journal on Communications, 2022, 43(3): 164-171.

图/表 6

图1

表1

表2

表3

图2

图3

参考文献 25

[1]	SALAKHUTDINOV R R , MNIH A . Probabilistic matrix factorization[C]// Proceedings of the 20th International Conference on Neural Information Processing Systems. New York:Curran Associates Inc, 2007: 1257-1264.
[2]	KOREN Y , BELL R , VOLINSKY C . Matrix factorization techniques for recommender systems[J]. Computer, 2009,42(8): 30-37.
[3]	任开旭, 王玉龙, 刘同存 ,等. 融合多维语义表示的概率矩阵分解模型[J]. 电子学报, 2019,47(9): 1848-1854.
	REN K X , WANG Y L , LIU T C ,et al. A probabilistic matrix factorization model based on multidimensional semantic representation learning[J]. Acta Electronica Sinica, 2019,47(9): 1848-1854.
[4]	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. New York:ACM Press, 2016: 233-240.
[5]	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. New York:ACM Press, 2017: 425-434.
[6]	LIU T C , LIAO J X , WANG Y L ,et al. Collaborative tensor-topic factorization model for personalized activity recommendation[J]. Multimedia Tools and Applications, 2019,78(12): 16923-16943.
[7]	顾亦然, 姚朱鹏, 杨海根 . 融合注意力胶囊的深度因子分解机模型[J]. 通信学报, 2021,42(10): 130-139.
	GU Y R , YAO Z P , YANG H G . Deep factorization machine model based on attention capsule[J]. Journal on Communications, 2021,42(10): 130-139.
[8]	CHEN C , ZHANG M , LIU Y Q ,et al. Neural attentional rating regression with review-level explanations[C]// Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW’18. New York:ACM Press, 2018: 1583-1592.
[9]	LIAO J X , LIU T C , YIN H Z ,et al. An integrated model based on deep multimodal and rank learning for point-of-interest recommendation[J]. World Wide Web, 2021,24(2): 631-655.
[10]	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. New York:ACM Press, 2013: 165-172.
[11]	BAO Y , FANG H , ZHANG J . TopicMF:simultaneously exploiting ratings and reviews for recommendation[C]// Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence. Palo Alto:AAAI Press, 2014: 2-8.
[12]	WANG H , WANG N Y , YEUNG D Y . Collaborative deep learning for recommender systems[C]// Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York:ACM Press, 2015: 1235-1244.
[13]	VINCENT P , LAROCHELLE H , LAJOIE I ,et al. Stacked denoising autoencoders:learning useful representations in a deep network with a local denoising criterion[J]. Journal of Machine Learning Research, 2010,11: 3371-3408.
[14]	LU Y C , DONG R H , SMYTH B . Coevolutionary recommendation model:mutual learning between ratings and reviews[C]// Proceedings of the 2018 World Wide Web Conference on World Wide Web WWW’18. New York:ACM Press, 2018: 773-782.
[15]	SEO S , HUANG J , YANG H ,et al. Interpretable convolutional neural networks with dual local and global attention for review rating prediction[C]// Proceedings of the Eleventh ACM Conference on Recommender Systems. New York:ACM Press, 2017: 297-305.
[16]	TAY Y , LUU A T , HUI S C . Multi-pointer co-attention networks for recommendation[C]// Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery ＆ Data Mining. New York:ACM Press, 2018: 2309-2318.
[17]	MA C , KANG P , WU B ,et al. Gated attentive-autoencoder for content-aware recommendation[C]// Proceedings of the Twelfth ACM International Conference on Web Search and Data Mining. New York:ACM Press, 2019: 519-527.
[18]	ZHOU J P , CHENG Z Y , PEREZ F ,et al. TAFA:two-headed attention fused autoencoder for context-aware recommendations[C]// Proceedings of the Fourteenth ACM Conference on Recommender Systems. New York:ACM Press, 2020: 338-347.
[19]	LIU H T , WU F Z , WANG W J ,et al. NRPA:neural recommendation with personalized attention[C]// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York:ACM Press, 2019: 1233-1236.
[20]	梁顺攀, 刘伟, 尤殿龙 ,等. 考虑评论质量的自注意力胶囊网络评分预测模型[J]. 电子与信息学报, 2021,43(12): 3451-3458.
	LIANG S P , LIU W , YOU D L ,et al. Self-attention capsule network rate prediction with review quality[J]. Journal of Electronics ＆ Information Technology, 2021,43(12): 3451-3458.
[21]	冯兴杰, 曾云泽 . 基于评分矩阵与评论文本的深度推荐模型[J]. 计算机学报, 2020,43(5): 884-900.
	FENG X J , ZENG Y Z . Joint deep modeling of rating matrix and reviews for recommendation[J]. Chinese Journal of Computers, 2020,43(5): 884-900.
[22]	李昆仑, 翟利娜, 赵佳耀 ,等. 融合信任关系与评论文本的矩阵分解推荐算法[J]. 小型微型计算机系统, 2021,42(2): 285-290.
	LI K L , ZHAI L N , ZHAO J Y ,et al. Matrix factorization recommendation algorithms by exploiting trust relationship and review text[J]. Journal of Chinese Computer Systems, 2021,42(2): 285-290.
[23]	GAO J Y , WANG X T , WANG Y S ,et al. Explainable recommendation through attentive multi-view learning[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto:AAAI Press, 2019: 3622-3629.
[24]	BERG R V D , KIPF T N , WELLING M . Graph convolutional matrix completion[J]. arXiv Preprint,arXiv:1706.02263, 2017.
[25]	GAO J Y , LIN Y , WANG Y S ,et al. Set-sequence-graph:a multi-view approach towards exploiting reviews for recommendation[C]// Proceedings of the 29th ACM International Conference on Information ＆ Knowledge Management. New York:ACM Press, 2020: 395-404.

数据集	用户数	商品数	评分数
MI	1 429	900	10 261
PLG	1 686	962	13 272
Auto	2 928	1 835	20 473
AIV	5 130	1 685	37 126

推荐模型	评分数据	评论文本	卷积神经网络	注意力机制	图卷积编码
PMF	3	—	—	—	—
LFM	3	—	—	—	—
ConvMF	3	3	3	—	—
DeepCoNN	—	3	3	—	—
NARRE	—	3	3	3	—
GC-MC	3	—	—	—	3
CRRG	3	3	3	3	3

推荐模型	MI	PLG	Auto	AIV
PMF	1.363 6	1.492 3	1.506 4	1.424 5
LFM	1.792 0	1.671 9	1.798 6	1.485 8
ConvMF	1.357 4	1.331 8	1.283 4	1.228 3
DeepCoNN	0.835 7	1.070 9	0.835 5	1.018 8
NARRE	0.804 2	1.025 2	0.788 5	0.961 2
GC-MC	1.129 9	1.292 4	1.024 3	1.151 5
CRRG	0.791 4	1.017 1	0.785 9	0.959 8

融合评论文本特征和评分图卷积表示的推荐模型

Recommendation model combining review’s feature and rating graph convolutional representation

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 25

相关文章 10

Metrics

推荐阅读 0

[1]	仲美玉, 吴培良, 窦燕, 刘毅, 孔令富. 基于中文语义-音韵信息的语音识别文本校对模型[J]. 通信学报, 2022, 43(11): 65-79.
[2]	陈卓, 朱淼, 杜军威. 基于多视角图神经网络的欺诈检测算法[J]. 通信学报, 2022, 43(11): 225-232.
[3]	王洪雁, 袁海. 基于骨骼及表观特征融合的动作识别方法[J]. 通信学报, 2022, 43(1): 138-148.
[4]	赵晓娟, 贾焰, 李爱平, 陈恺. 基于层级注意力机制的链接预测模型研究[J]. 通信学报, 2021, 42(3): 36-44.
[5]	王桐, 高山, 龚慧雯, 孙博. 基于分时MDP的出租车载客预测推荐技术研究[J]. 通信学报, 2021, 42(2): 37-51.
[6]	顾亦然, 姚朱鹏, 杨海根. 融合注意力胶囊的深度因子分解机模型[J]. 通信学报, 2021, 42(10): 130-139.
[7]	郭璠, 张泳祥, 唐琎, 李伟清. YOLOv3-A：基于注意力机制的交通标志检测网络[J]. 通信学报, 2021, 42(1): 87-99.
[8]	李琳辉,周彬,连静,周雅夫. 基于社会注意力机制的行人轨迹预测方法研究[J]. 通信学报, 2020, 41(6): 175-183.
[9]	王竹荣, 薛伟, 牛亚邦, 崔颖安, 孙钦东, 黑新宏. 基于注意力机制的泊位占有率预测模型研究[J]. 通信学报, 2020, 41(12): 182-192.
[10]	琚春华,鲍福光. 基于情境和主体特征融入性的多维度个性化推荐模型研究[J]. 通信学报, 2012, 33(Z1): 17-27.