Hybrid sentiment analysis with textual and interactive information

作者：

Highlights：

• Incorporate textual and interactive information for sentiment analysis.

• Learn discriminative features with a novel ranking graph convolutional network.

• Preserve edge semantics during the graph convolutional propagation.

• Demonstrate the advantages of the proposed method with extensive experiments.

摘要

•Incorporate textual and interactive information for sentiment analysis.•Learn discriminative features with a novel ranking graph convolutional network.•Preserve edge semantics during the graph convolutional propagation.•Demonstrate the advantages of the proposed method with extensive experiments.

论文关键词：Interactive data learning,Discriminative representation learning

论文评审过程：Received 29 March 2022, Revised 6 September 2022, Accepted 1 October 2022, Available online 9 October 2022, Version of Record 17 October 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118960