Hierarchical and lateral multiple timescales gated recurrent units with pre-trained encoder for long text classification
作者：
Highlights：
• Performance of current text classifiers degrade with longer input sequences.
• The proposed model can classify texts of diverse input lengths.
• A hierarchical and lateral architecture is proposed to enhance the performance.
• The model uses rich features extracted by pre-trained bidirectional encoders.
• Our model outperforms existing models on various long text classification datasets.
摘要
•Performance of current text classifiers degrade with longer input sequences.•The proposed model can classify texts of diverse input lengths.•A hierarchical and lateral architecture is proposed to enhance the performance.•The model uses rich features extracted by pre-trained bidirectional encoders.•Our model outperforms existing models on various long text classification datasets.
论文关键词：Text classification,Multiple timescale,Temporal hierarchy,BERT,Pre-trained encoder
论文评审过程：Received 26 February 2020, Revised 28 July 2020, Accepted 16 August 2020, Available online 21 August 2020, Version of Record 3 September 2020.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.113898
