采用注意力门控卷积网络模型的目标情感分析

doi:10.19665/j.issn1001-2400.2019.06.005

西安电子科技大学学报 ›› 2019, Vol. 46 ›› Issue (6): 30-36.doi: 10.19665/j.issn1001-2400.2019.06.005

采用注意力门控卷积网络模型的目标情感分析

曹卫东,李嘉琪(),王怀超

中国民航大学计算机科学与技术学院, 天津 300300

收稿日期:2019-06-24 出版日期:2019-12-20 发布日期:2019-12-21
通讯作者: 李嘉琪
作者简介:曹卫东(1964—),女,副教授,博士,E-mail:2471891478@qq.com
基金资助:
民航科技创新重大专项(MHRD20160109);民航安全能力项目(TRSA201803)

Analysis of targeted sentiment by the attention gated convolutional network model

CAO Weidong,LI Jiaqi(),WANG Huaichao

School of Computer Science and Technology, Civil Aviation University of China, Tianjin 300300, China

Received:2019-06-24 Online:2019-12-20 Published:2019-12-21
Contact: Jiaqi LI

摘要/Abstract

摘要：

针对传统目标情感分析采用循环神经网络模型导致训练时间长且其他替代模型未能使得上下文和目标词之间实现良好交互等问题,提出了一种用于目标情感分析的注意力门控卷积网络模型。该模型首先将上下文和目标词通过多头注意力机制加强上下文和目标词之间的交互;其次采用门控卷积机制进一步提取关于目标词的情感特征;最后通过Softmax分类器将情感特征进行分类,输出情感极性。实验结果显示,与循环神经网络模型中准确率最高的循环注意力网络模型相比,在SemEval 2014任务四的餐厅和笔记本电脑数据集上的准确率分别提高了1.29%和0.12%;与循环神经网络模型中收敛速度较快的基于注意力的长短期记忆网络模型相比,收敛时间下降了约29.17s。

关键词: 情感分析, 循环神经网络, 多头注意力机制, 门控卷积机制

Abstract:

The recurrent neural networks are used for traditional targeted sentiment analysis and usually lead to a long training time. And other alternative models are unable to make a good interaction between context and target words. An attention gated convolutional network model for targeted sentiment analysis is proposed. First, context and target words are processed by the multiple attention mechanism to enhance their interactions. Second, the gated convolution mechanism is used to selectively generate emotional features. Finally, the emotional features are classified by the Softmax classifier to output the emotional polarity. Experimental results show that compared with the Recurrent Attention Network model, which has the highest accuracy rate in the recurrent neural network models, the proposed model improves the accuracy rate by 1.29% and 0.12% respectively on the Restaurant and Laptop datasets of SemEval 2014 Task4. Compared with the Attention-based Long Short-Term Memory Network model, which has a faster convergence rate in the recurrent neural network model, the convergence time is reduced by 29.17 s.

Key words: sentiment analysis, recurrent neural networks, multiple attention mechanism, gated convolution mechanism

中图分类号:

TP391

曹卫东,李嘉琪,王怀超. 采用注意力门控卷积网络模型的目标情感分析[J]. 西安电子科技大学学报, 2019, 46(6): 30-36.

CAO Weidong,LI Jiaqi,WANG Huaichao. Analysis of targeted sentiment by the attention gated convolutional network model[J]. Journal of Xidian University, 2019, 46(6): 30-36.

图/表 6

图1

表1

表2

表3

表4

图2

参考文献 14

[1]	WANG S, MAZUMDER S, LIU B , et al. Target-sensitive Memory Networks for Aspect Sentiment Classification[C]//Proceedings of the 2018 56th Annual Meeting of the Association for Computational Linguistics. New York: ACL, 2018: 957-967.
[2]	MA D H, Li S J, WANG H F . Joint Learning for Targeted Sentiment Analysis[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2018: 4737-4742.
[3]	LIU F, COHN T, BALDWIN T . Recurrent Entity Networks with Delayed Memory Update for Targeted Aspect-based Sentiment Analysis[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2018: 278-283.
[4]	CHEN P, SUN Z, BING L , et al. Recurrent Attention Network on Memory for Aspect Sentiment Analysis[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2017: 452-461.
[5]	TANG D, QIN B, LIU T . Aspect Level Sentiment Classification with Deep Memory Network[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2016: 214-224.
[6]	WANG Y, HUANG M, ZHU X , et al. Attention-based LSTM for Aspect-level Sentiment Classification[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2016: 606-615.
[7]	MA D H, LI S J, ZHANG X D , et al. Interactive Attention Networks for Aspect-level Sentiment Classification[C/OL]// [2019-05-26]. https://arxiv.org/pdf/1709.00893.pdf.
[8]	MISHRA A, DEY K, BHATTACHARYYA P . Learning Cognitive Features from Gaze Data for Sentiment and Sarcasm Classification Using Convolutional Neural Network[C]//Proceedings of the 2017 55th Annual Meeting of the Association for Computational Linguistics. New York: ACL, 2017: 377-387.
[9]	XUE W, LI T . Aspect Based Sentiment Analysis with Gated Convolutional Networks[C]//Proceedings of the 2018 56th Annual Meeting of the Association for Computational Linguistics. New York: ACL, 2018: 2514-2523.
[10]	LI X, BING L, LAM W , et al. Transformation Networks for Target-oriented Sentiment Classification[C]//Proceedings of the 2018 56th Annual Meeting of the Association for Computational Linguistics. New York: ACL, 2018: 946-956.
[11]	PENNINGTON J, SOCHER R, MANNING C D . Glove: Global Vectors for Word Representation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. New York: ACL, 2014: 1532-1543.
[12]	Song Y W, Wang J H, Jiang T , et al. Attentional Encoder Network for Targeted Sentiment Classification[J/OL].[2019-05-26]. https://arxiv.org/pdf/1902.09314.pdf.
[13]	刘道华, 崔玉爽, 赵岩松 , 等. 一种改进卷积神经网络的教学图像检索方法[J]. 西安电子科技大学学报, 2019,46(3):52-58.
	LIU Daohua, CUI Yushuang, ZHAO Yansong , et al. Method for Retrieving the Teaching Image Based on the Improved Convolutional Neural Network[J]. Journal of Xidian University, 2019,46(3) : 52-58.
[14]	陈珂, 梁斌, 柯文德 , 等. 基于多通道卷积神经网络的中文微博情感分析[J]. 计算机研究与发展, 2018,55(5 ): 945-957.
	CHEN Ke, LIANG Bin, KE Wende , et al. Chinese Micro-blog Sentiment Analysis Based on Multi-channels Convolutional Neural Networks[J]. Journal of Computer Research and Development, 2018,55(5) : 945-957.

数据集	积极		中性		消极
数据集	训练集	测试集	训练集	测试集	训练集	测试集
餐厅	2 164	728	637	196	807	196
笔记本电脑	994	341	464	169	870	128

超参数	餐厅参数值	笔记本电脑参数值
词向量维度	300	300
卷积核个数	100	100
卷积核尺寸	3, 4, 5	3, 4, 5
学习率	0.005	0.002
随机失活	0.2	0.1
批大小	128	128
最大迭代次数	20	20
优化器	AdaGrad	Adam

网络模型		餐厅准确率	笔记本电脑准确率
循环神经网络模型	TD-LSTM	0.756 3	0.681 3
	ATAE-LSTM	0.772 0	0.687 0
	IAN	0.786 0	0.721 0
	RAM	0.802 3	0.744 9
非循环神经网络模型	MemNet	0.781 6	0.703 3
	GCAE	0.794 6	0.705 3
	AEN-GloVe	0.809 8	0.735 1
文中模型	AGCN	0.815 2	0.746 1

网络模型		收敛时间/s
循环神经网络模型	TD-LSTM	10.50
	ATAE-LSTM	55.78
	IAN	131.87
	RAM	80.93
非循环神经网络模型	MemNet	22.94
	GCAE	6.48
	AEN-GloVe	65.61
文中模型	AGCN	26.61

采用注意力门控卷积网络模型的目标情感分析

Analysis of targeted sentiment by the attention gated convolutional network model

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 14

相关文章 6

Metrics

本文评价

推荐阅读 10

[1]	李源,崔玉爽,王伟. 一种基于字词双通道网络的文本情感分析方法[J]. 西安电子科技大学学报, 2021, 48(6): 179-186.
[2]	徐彬,张永顺,张秦,王富平,郑桂妹. 采用乘性RNN的雷达HRRP目标识别[J]. 西安电子科技大学学报, 2021, 48(2): 49-54.
[3]	张璐,孙蓉,刘景伟. 分布式存储中的克隆piggybacking框架[J]. 西安电子科技大学学报, 2020, 47(6): 139-147.
[4]	胡建伟,赵伟,崔艳鹏,崔俊洁. 一种改进ASTNN网络的PHP代码漏洞挖掘方法[J]. 西安电子科技大学学报, 2020, 47(6): 164-173.
[5]	张志昌,张治满,张珍文. 融合局部语义和全局结构信息的健康问句分类[J]. 西安电子科技大学学报, 2020, 47(2): 9-15.
[6]	马书磊,张国宾,焦阳,石光明. 一种改进的全局注意机制图像描述方法[J]. 西安电子科技大学学报, 2019, 46(2): 17-22.