Hierarchical attention and transformers for automatic movie rating

作者：

Highlights：

• A novel Hierarchical RNN and Transformer architecture well suited for movie scripts.

• Novel and effective strategies to analyze movie scripts as large text sequences.

• A new visualization strategy to improve the interpretability in transformers.

摘要

•A novel Hierarchical RNN and Transformer architecture well suited for movie scripts.•Novel and effective strategies to analyze movie scripts as large text sequences.•A new visualization strategy to improve the interpretability in transformers.

论文关键词：Hierarchical transformers,Movie rating prediction,Text classification,Transformer visualization

论文评审过程：Received 4 November 2021, Revised 14 June 2022, Accepted 10 July 2022, Available online 18 July 2022, Version of Record 3 August 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118164