（八十四）：A Hierarchical Network for Abstractive Meeting Summarization with Cross-Domain Pretraining

• Abstract
• 1. Introduction
• 2. Problem Formulation
• 3. Method
• • 3.1 Encoder
  • 3.1.1 Role Vector
  • 3.1.2 Hierarchical Transformer
  • 3.2 Decoder
  • 3.3 Pretraining
• 4. Experiment
• 5. Evaluation Results
• 7. Conclusions

• 出处： EMNLP (Findings) 2020: 194-203
• 代码：https://github.com/microsoft/HMNet</