Loss-Aware Curriculum Learning for Chinese Grammatical Error Correction

TL;DR

This paper introduces a loss-aware curriculum learning framework for Chinese grammatical error correction that orders training samples by difficulty and adjusts the loss function to improve model performance.

Contribution

It proposes a novel multi-granularity curriculum learning approach that considers correction difficulty and dynamically regulates the loss function during training.

Findings

Significant performance improvements on multiple datasets.

Effective sample difficulty estimation method.

Enhanced model training stability and accuracy.

Abstract

Chinese grammatical error correction (CGEC) aims to detect and correct errors in the input Chinese sentences. Recently, Pre-trained Language Models (PLMS) have been employed to improve the performance. However, current approaches ignore that correction difficulty varies across different instances and treat these samples equally, enhancing the challenge of model learning. To address this problem, we propose a multi-granularity Curriculum Learning (CL) framework. Specifically, we first calculate the correction difficulty of these samples and feed them into the model from easy to hard batch by batch. Then Instance-Level CL is employed to help the model optimize in the appropriate direction automatically by regulating the loss function. Extensive experimental results and comprehensive analyses of various datasets prove the effectiveness of our method.