Data Weighted Training Strategies for Grammatical Error Correction

TL;DR

This paper explores data weighting strategies for grammatical error correction, demonstrating that incorporating delta-log-perplexity scores into training improves model performance and achieves state-of-the-art results.

Contribution

It introduces the use of delta-log-perplexity for data weighting in GEC training, providing empirical insights and demonstrating improved results.

Findings

Scored data with delta-log-perplexity enhances GEC model performance.

Incorporating data scores into training schedules leads to state-of-the-art results.

Empirical analysis clarifies the role of delta-log-perplexity in GEC training.

Abstract

Recent progress in the task of Grammatical Error Correction (GEC) has been driven by addressing data sparsity, both through new methods for generating large and noisy pretraining data and through the publication of small and higher-quality finetuning data in the BEA-2019 shared task. Building upon recent work in Neural Machine Translation (NMT), we make use of both kinds of data by deriving example-level scores on our large pretraining data based on a smaller, higher-quality dataset. In this work, we perform an empirical study to discover how to best incorporate delta-log-perplexity, a type of example scoring, into a training schedule for GEC. In doing so, we perform experiments that shed light on the function and applicability of delta-log-perplexity. Models trained on scored data achieve state-of-the-art results on common GEC test sets.