Multi-label legal document classification: A deep learning-based approach with label-attention and domain-specific pre-training

作者：

Highlights：

• A novel legal procedural posture dataset (50K cases) for multi-label classification

• Deep learning architecture: LM pre-training, label embedding and multi-task learning

• Substantially better performance than four recent methods on two large-scale datasets

摘要

•A novel legal procedural posture dataset (50K cases) for multi-label classification•Deep learning architecture: LM pre-training, label embedding and multi-task learning•Substantially better performance than four recent methods on two large-scale datasets

论文关键词：Multi-label classification,Procedural posture,Label-attended document classification,Domain-specific pre-training

论文评审过程：Received 30 October 2020, Accepted 4 January 2021, Available online 18 January 2021, Version of Record 9 February 2022.

论文官网地址：https://doi.org/10.1016/j.is.2021.101718