Grammatical Error Correction with Neural Reinforcement Learning

Abstract

We propose a neural encoder-decoder model with reinforcement learning (NRL) for grammatical error correction (GEC). Unlike conventional maximum likelihood estimation (MLE), the model directly optimizes towards an objective that considers a sentence-level, task-specific evaluation metric, avoiding the exposure bias issue in MLE. We demonstrate that NRL outperforms MLE both in human and automated evaluation metrics, achieving the state-of-the-art on a fluency-oriented GEC corpus.

References in corpus (3)

• Adam: A Method for Stochastic Optimization
• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Sequence Level Training with Recurrent Neural Networks

Cited by in corpus (4)

• On the Weaknesses of Reinforcement Learning for Neural Machine Translation
• ENCORE: Ensemble Learning using Convolution Neural Machine Translation for Automatic Program Repair
• Reference-less Measure of Faithfulness for Grammatical Error Correction
• Cross-Corpora Evaluation and Analysis of Grammatical Error Correction Models --- Is Single-Corpus Evaluation Enough?