Bi-graph attention network for aspect category sentiment classification

作者：

Highlights：

• Constructing two graphs to gain the sequential and syntactic information.

• Utilizing intra-graph and inter-graph transfer to encode sentence representations.

• Using the retrieval-based attention to reduce the noise caused by unrelated words.

• The experimental results demonstrate the state-of-the-art performance of BiGAT.

摘要

•Constructing two graphs to gain the sequential and syntactic information.•Utilizing intra-graph and inter-graph transfer to encode sentence representations.•Using the retrieval-based attention to reduce the noise caused by unrelated words.•The experimental results demonstrate the state-of-the-art performance of BiGAT.

论文关键词：Aspect category sentiment classification,Graph attention network,Biaffine module,Attention mechanism

论文评审过程：Received 16 May 2022, Revised 29 September 2022, Accepted 30 September 2022, Available online 7 October 2022, Version of Record 19 October 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.109972