基于句法结构与语义分析的方面级情感分析方法
Aspect-level Sentiment Analysis Method Based on Syntactic Structure and Semantic Analysis
周慧慧 1王善勤1
作者信息
- 1. 滁州职业技术学院 信息工程学院,安徽省 滁州 239000
- 折叠
摘要
为有效利用句子中的句法以及语义特征,本文提出基于句法结构与语义分析的方面级情感分析模型(Aspect-level sentiment analysis method based on syntactic structure and semantic analysis,LCF-Sync).具体来说,LCF-Sync根据句子中单词与方面之间的距离获取单词与方面的词法特征,并构建单词节点之间的句法依赖树获取句法特征;LCF-Sync利用多头自注意力机制获取句子中单词之间的语义特征,同时融合局部特征与全局特征进行情感分析预测.本文在三个基准数据集上进行大量实验,实验结果表明本文提出的模型优于基线方法.
Abstract
To effectively utilize the syntactic and semantic features in sentences,this paper proposes anaspect-level sentiment analysis method based on syntactic structure and semantic analysis (LCF-Sync). Specifically,LCF-Sync obtains lexical features of words and aspects based on the distance between them in a sentence,and constructs a syntactic dependency tree between nodes to obtain syntactic features.LCF-Sync utilizes a multi-head self-attention mechanism to obtain semantic features between words in sentences,while fusing local and global features for sentiment analysis and prediction.This paper has conducted extensive experiments on three benchmark datasets,and the experimental results show that the proposed model outperforms the baseline method.
关键词
情感分析/注意力机制/句法结构/语义特征Key words
sentiment analysis/attention mechanism/syntactic structure/semantic features引用本文复制引用
基金项目
安徽省高校自然科学研究重点项目(2023AH053091)
安徽省高校自然科学研究重点项目(2023AH053088)
滁州职业技术学院自然科学研究重点项目(ZKZ-2022-02)
安徽省高等学校省级质量工程项目(2022JNDS043)
出版年
2024
NETL
NSTL
万方数据