基于知识增强与多层注意力机制的方面级情感分析方法  

Aspect-Based Sentiment Analysis Method Based on Knowledge Enhancement and Multi-Layer Attention Mechanism

作　　者：黄礼东 胡慧君[1,2,3] 陈佳逸 刘茂福 HUANG Lidong;HU Hujun;CHEN Jiayi;LIU Maofu(College of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China;Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System,Wuhan 430065,Hubei,China;Big Data Science and Engineering Research Institute,Wuhan University of Science and Technology,Wuhan 430081,Hubei,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,湖北武汉430065 [2]智能信息处理与实时工业系统湖北省重点实验室,湖北武汉430065 [3]武汉科技大学大数据科学与工程研究院,湖北武汉430081

出　　处：《武汉大学学报（理学版）》2024年第4期473-481,共9页Journal of Wuhan University:Natural Science Edition

基　　金：“十四五”湖北省优势特色学科群项目(2023D0302);湖北省大学生创新创业训练项目(S202310488126)。

摘　　要：方面级文本情感分析旨在分析长句子中不同方面各自对应的情感极性。目前,较新的神经网络模型着重考虑了文本中方面词提取对情感预测的作用,但忽略了文本信息歧义和生僻等现实文本缺陷,导致外部知识和方面词信息之间的互补性还未得到充分的研究,模型性能仍有优化空间。因此,提出一种基于知识增强与多层注意力机制的方法,通过双向长短期记忆网络存储文本表示与辅助信息,利用图卷积神经网络获得句法信息并用知识增强来提取方面词的归类表示,引入多层注意力机制将融合后特征输入全局平均池化层,提高精准度。实验表明,在SemEval任务和Twitter数据集上,所提方法性能超过基线系统,获得了较好的实验结果。Aspect-level sentiment analysis aims to analyze the corresponding sentimental polarity of different aspects in long sentences.Currently,newer neural network models primarily concentrate on the impact of aspect term extraction on sentiment prediction but overlook real text defects such as ambiguity and lack of textual information.This oversight leads to inadequate exploration of the complementarity between external knowledge and aspect term information,leaving room for further optimization of model performance.To solve this problem,this paper proposes a method based on knowledge enhancement and a multi-layer attention mechanism.The model stores text representation and auxiliary information through a bi-directional long short-term memory network,obtains syntax information using a graph convolutional network,and extracts categorical representation of aspect terms by knowledge enhancement.This means a multi-layer attention mechanism is introduced to input the fused features into the global average pooling layer to improve accuracy.On the SemEval task and Twitter dataset,the proposed model outperforms the baseline and achieves good experimental results.

关 键 词：方面级情感分析 知识增强 图卷积神经网络 多层注意力机制 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP183[自动化与计算机技术—计算机科学与技术]