计算机应用研究2025，Vol.42Issue(9)：2669-2675,7.DOI:10.19734/j.issn.1001-3695.2025.03.0039
基于句法依存增强和方面语义聚焦的方面级情感分析
Aspect-based sentiment analysis based on syntactic dependency enhancement and aspect semantic focus
摘要
Abstract
Most existing aspect-based sentiment analysis studies use dependency trees-based graph neural networks to build models,but ignore the numerous irrelevant dependencies contained in the original dependency tree and lack the extraction of semantic features for specific aspects.This paper proposed a novel dual-channel graph convolutional network model(SADGCN),which relied on syntactic dependency enhancement and aspect semantic focus to improve the prediction accuracy of sentiment analysis.The proposed scheme consists of two main components:a syntactic enhancement module and a semantic enhancement module.For the syntactic enhancement module,it proposed an efficient aspect-oriented word selection method to reshape the syntactic dependency tree,thereby more accurately capturing syntactic information highly relevant to the target as-pect.The semantic enhancement module designed an aspect-focused attention mechanism integrated with the global self-atten-tion mechanism,which facilitated the learning process of semantic feature representations of specific aspects.Finally,it fused the information learned from both modules to predict sentiment polarity.Experimental results on three benchmark datasets demonstrate that the proposed model achieves better performance compared to the comparison model.关键词
方面级情感分析/句法结构/语义特征/图卷积网络/注意力机制Key words
aspect-based sentiment analysis(ABSA)/syntactic structure/semantic features/graph convolutional network(GCN)/attention mechanism分类
信息技术与安全科学引用本文复制引用
王一力,陈浩文,袁程胜..基于句法依存增强和方面语义聚焦的方面级情感分析[J].计算机应用研究,2025,42(9):2669-2675,7.基金项目
国家自然科学基金企业联合基金重点项目(U23B2023) （U23B2023）
国家自然科学基金青年项目(62102189) （62102189）
