Neural Language Correction with Character-Based Attention

TL;DR

This paper introduces a character-based neural encoder-decoder model with attention for language correction, effectively handling orthographic errors and improving performance on learner text datasets.

Contribution

The paper presents a novel character-level neural correction model with attention, outperforming previous methods and demonstrating the benefit of training on synthesized errors.

Findings

Achieved state-of-the-art F0.5 score on CoNLL 2014 dataset.

Character-level model handles out-of-vocabulary and orthographic errors.

Training with synthesized errors improves correction performance.

Abstract

Natural language correction has the potential to help language learners improve their writing skills. While approaches with separate classifiers for different error types have high precision, they do not flexibly handle errors such as redundancy or non-idiomatic phrasing. On the other hand, word and phrase-based machine translation methods are not designed to cope with orthographic errors, and have recently been outpaced by neural models. Motivated by these issues, we present a neural network-based approach to language correction. The core component of our method is an encoder-decoder recurrent neural network with an attention mechanism. By operating at the character level, the network avoids the problem of out-of-vocabulary words. We illustrate the flexibility of our approach on dataset of noisy, user-generated text collected from an English learner forum. When combined with a language model, our method achieves a state-of-the-art $F_{0.5}$-score on the CoNLL 2014 Shared Task. We further demonstrate that training the network on additional data with synthesized errors can improve performance.