Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing

doi:10.3390/electronics13101993

	Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing
	Xu, Erfeng 1,2; Zhu, Junwu 1; Zhang, Luchen 3; Wang, Yi 1,2; Lin, Wei 1,2
	2024-05-01
发表期刊	ELECTRONICS
卷号	13 期号:10 页码:16
摘要	Aspect-level sentiment analysis is used to predict the sentiment polarity of a specific aspect in a sentence. However, most current research cannot fully utilize semantic information, and the models lack robustness. Therefore, this article proposes a model for aspect-level sentiment analysis based on a combination of adversarial training and dependency syntax analysis. First, BERT is used to transform word vectors and construct adjacency matrices with dependency syntactic relationships to better extract semantic dependency relationships and features between sentence components. A multi-head attention mechanism is used to fuse the features of the two parts, simultaneously perform adversarial training on the BERT embedding layer to enhance model robustness, and, finally, to predict emotional polarity. The model was tested on the SemEval 2014 Task 4 dataset. The experimental results showed that, compared with the baseline model, the model achieved significant performance improvement after incorporating adversarial training and dependency syntax relationships.
关键词	multi head attention mechanism dependency syntactic relationships adjacency matrix adversarial training
DOI	10.3390/electronics13101993
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China
WOS研究方向	Computer Science ; Engineering ; Physics
WOS类目	Computer Science, Information Systems ; Engineering, Electrical & Electronic ; Physics, Applied
WOS记录号	WOS:001232935000001
出版者	MDPI
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40045
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Zhu, Junwu; Zhang, Luchen
作者单位	1.Yangzhou Univ, Sch Informat Engn, Yangzhou 225127, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 3.Natl Comp Network Emergency Response Tech Team, Coordinat Ctr China, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Xu, Erfeng,Zhu, Junwu,Zhang, Luchen,et al. Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing[J]. ELECTRONICS,2024,13(10):16.
APA	Xu, Erfeng,Zhu, Junwu,Zhang, Luchen,Wang, Yi,&Lin, Wei.(2024).Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing.ELECTRONICS,13(10),16.
MLA	Xu, Erfeng,et al."Research on Aspect-Level Sentiment Analysis Based on Adversarial Training and Dependency Parsing".ELECTRONICS 13.10(2024):16.