Bidirectional Edge-Enhanced Graph Convolutional Networks for Aspect-based Sentiment Classification

Du, Jinyang; Zhang, Yin; Yue, Binglei; Chen, Min

doi:10.1109/compsac51774.2021.00238

Cited by 3 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lu [18] designed a gated mechanism based on Bi-LSTM to guide the encoding of sentiment information related to aspects, followed by a GCN to capture long-range dependencies of words. After concatenating dependency relations and their two sides words, Du [19] employed a GCN and a multi-head attention mechanism to capture aspect-related sentiment information. Wang [5] reconstructed dependency trees by taking aspects as roots and applied BERT and graph attention networks (GATs) to explicitly encode dependency relations.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Contextual Graph Attention Network for Aspect-Level Sentiment Classification

et al. 2022

View full text Add to dashboard Cite

Aspect-level sentiment classification aims to predict the sentiment polarities towards the target aspects given in sentences. To address the issues of insufficient semantic information extraction and high computational complexity of attention mechanisms in existing aspect-level sentiment classification models based on deep learning, a contextual graph attention network (CGAT) is proposed. The proposed model adopts two graph attention networks to aggregate syntactic structure information into target aspects and employs a contextual attention network to extract semantic information in sentence-aspect sequences, aiming to generate aspect-sensitive text features. In addition, a syntactic attention mechanism based on syntactic relative distance is proposed, and the Gaussian function is cleverly introduced as a syntactic weight function, which can reduce computational complexities and effectively highlight the words related to aspects in syntax. Experiments on three public sentiment datasets show that the proposed model can make better use of semantic information and syntactic structure information to improve the accuracy of sentiment classification.

show abstract

Section: Related Workmentioning

confidence: 99%

“…ASGCN [4] employs a GCN to extract syntactic structure information. BE-GCN [19] combines words and their dependencies and adopts a GCN and a multi-head attention mechanism to capture aspect-related sentiment information.…”

Section: Baselinesmentioning

confidence: 99%