RobustGEC: Robust Grammatical Error Correction Against Subtle Context Perturbation

TL;DR

RobustGEC introduces a benchmark to evaluate the context robustness of GEC systems and reveals that current models lack sufficient stability against input perturbations, proposing a simple method to improve robustness.

Contribution

The paper presents RobustGEC, a new benchmark with 5,000 cases for testing GEC system robustness and proposes a method to enhance their stability against context perturbations.

Findings

Current GEC systems lack robustness against context perturbations.

RobustGEC benchmark effectively evaluates GEC system stability.

Proposed method improves GEC system robustness.

Abstract

Grammatical Error Correction (GEC) systems play a vital role in assisting people with their daily writing tasks. However, users may sometimes come across a GEC system that initially performs well but fails to correct errors when the inputs are slightly modified. To ensure an ideal user experience, a reliable GEC system should have the ability to provide consistent and accurate suggestions when encountering irrelevant context perturbations, which we refer to as context robustness. In this paper, we introduce RobustGEC, a benchmark designed to evaluate the context robustness of GEC systems. RobustGEC comprises 5,000 GEC cases, each with one original error-correct sentence pair and five variants carefully devised by human annotators. Utilizing RobustGEC, we reveal that state-of-the-art GEC systems still lack sufficient robustness against context perturbations. In addition, we propose a simple yet effective method for remitting this issue.