Hierarchical Curriculum Learning for AMR Parsing
This work addresses a specific bottleneck in AMR parsing for NLP researchers, offering an incremental improvement through curriculum learning.
The paper tackles the gap between flat training objectives and hierarchical structures in AMR parsing by proposing a Hierarchical Curriculum Learning framework, which improves model generalization and achieves strong results on benchmarks like AMR2.0 and AMR3.0.
Abstract Meaning Representation (AMR) parsing aims to translate sentences to semantic representation with a hierarchical structure, and is recently empowered by pretrained sequence-to-sequence models. However, there exists a gap between their flat training objective (i.e., equally treats all output tokens) and the hierarchical AMR structure, which limits the model generalization. To bridge this gap, we propose a Hierarchical Curriculum Learning (HCL) framework with Structure-level (SC) and Instance-level Curricula (IC). SC switches progressively from core to detail AMR semantic elements while IC transits from structure-simple to -complex AMR instances during training. Through these two warming-up processes, HCL reduces the difficulty of learning complex structures, thus the flat model can better adapt to the AMR hierarchy. Extensive experiments on AMR2.0, AMR3.0, structure-complex and out-of-distribution situations verify the effectiveness of HCL.