文献类型：期刊文献

英文题名：Aspect-Level Sentiment Analysis Based on Aspect Tree and Syntactic Matrix

作者：Cao, Hao[1]; Guo, Weibin[1]; Cheng, Jiahui[1]

机构：[1] School of Information Science and Engineering, East China University of Science and Technology, Shanghai, 200237, China

年份：2024

起止页码：34

外文期刊名：2024 IEEE 2nd International Conference on Image Processing and Computer Applications, ICIPCA 2024

收录：EI(收录号：20244517315652)

语种：英文

外文关键词：Convolutional neural networks - Matrix algebra - Network theory (graphs) - Syntactics - Trees (mathematics)

摘要：Graph convolutional network models based on syntactic dependency trees have been shown to be useful for aspect-level sentiment analysis, but the existing models do not make full use of syntactic structure information. To solve this problem, this paper proposes an aspect-level sentiment analysis method based on aspect tree and syntactic matrix. The original syntactic dependency tree is reshaped into an aspect-oriented tree, retaining the dependency types that are directly related to the aspect. At the same time, in order to effectively integrate syntactic and semantic information, syntactic matrices are used to equip attention matrices. Experimental results show that the new graph convolutional network model based on aspect tree and syntactic matrix shows better analysis performance on multiple public datasets than the baseline method, and the accuracy of sentiment classification on Laptop, Restaurant and Twitter is as high as 79.91%, 85.97% and 76.07%, respectively, indicating the effectiveness of the proposed method. ? 2024 IEEE.