Adapting Sequence Models for Sentence Correction

TL;DR

This paper demonstrates that character-based sequence models with diff-based output modeling outperform traditional phrase-based models in sentence correction, achieving higher accuracy with less data.

Contribution

It introduces diff-based output modeling for sequence-to-sequence sentence correction, showing improved performance over existing models and methods.

Findings

Character-based models outperform word-based models.

Diff modeling improves effectiveness over standard approaches.

Achieves 6 M2 points improvement over phrase-based SMT.

Abstract

In a controlled experiment of sequence-to-sequence approaches for the task of sentence correction, we find that character-based models are generally more effective than word-based models and models that encode subword information via convolutions, and that modeling the output data as a series of diffs improves effectiveness over standard approaches. Our strongest sequence-to-sequence model improves over our strongest phrase-based statistical machine translation model, with access to the same data, by 6 M2 (0.5 GLEU) points. Additionally, in the data environment of the standard CoNLL-2014 setup, we demonstrate that modeling (and tuning against) diffs yields similar or better M2 scores with simpler models and/or significantly less data than previous sequence-to-sequence approaches.