# Reaching Human-level Performance in Automatic Grammatical Error Correction: An Empirical Study

arXiv: https://arxiv.org/abs/1807.01270 · HTML: https://mlsanity.com/paper/1807.01270/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Jul 2018.

## Built on by (most cited first, 22 of 22)

- [PEER: A Collaborative Language Model](https://mlsanity.com/paper/2208.11663/) (2022)
- [Synthetic Data Generation for Grammatical Error Correction with Tagged Corruption Models](https://mlsanity.com/paper/2105.13318/) (2021)
- [Instantaneous Grammatical Error Correction with Shallow Aggressive Decoding](https://mlsanity.com/paper/2106.04970/) (2021)
- [EditEval: An Instruction-Based Benchmark for Text Improvements](https://mlsanity.com/paper/2209.13331/) (2022)
- [Towards Automated Document Revision: Grammatical Error Correction, Fluency Edits, and Beyond](https://mlsanity.com/paper/2205.11484/) (2022)
- [Tail-to-Tail Non-Autoregressive Sequence Prediction for Chinese Grammatical Error Correction](https://mlsanity.com/paper/2106.01609/) (2021)
- [A Unified Strategy for Multilingual Grammatical Error Correction with Pre-trained Cross-Lingual Language Model](https://mlsanity.com/paper/2201.10707/) (2022)
- [Misspelling Correction with Pre-trained Contextual Language Model](https://mlsanity.com/paper/2101.03204/) (2021)
- [Comparison of Grammatical Error Correction Using Back-Translation Models](https://mlsanity.com/paper/2104.07848/) (2021)
- [uChecker: Masked Pretrained Language Models as Unsupervised Chinese Spelling Checkers](https://mlsanity.com/paper/2209.07068/) (2022)
- [Understanding Model Robustness to User-generated Noisy Texts](https://mlsanity.com/paper/2110.07428/) (2021)
- [Adapting LLMs for Minimal-edit Grammatical Error Correction](https://mlsanity.com/paper/2506.13148/) (2025)
- [Jointly Learning to Repair Code and Generate Commit Message](https://mlsanity.com/paper/2109.12296/) (2021)
- [Variation and generality in encoding of syntactic anomaly information in sentence embeddings](https://mlsanity.com/paper/2111.06644/) (2021)
- [Towards Lithuanian grammatical error correction](https://mlsanity.com/paper/2203.09963/) (2022)
- [An Alignment-Agnostic Model for Chinese Text Error Correction](https://mlsanity.com/paper/2104.07190/) (2021)
- [Improvement in Machine Translation with Generative Adversarial Networks](https://mlsanity.com/paper/2111.15166/) (2021)
- [Pretraining Chinese BERT for Detecting Word Insertion and Deletion Errors](https://mlsanity.com/paper/2204.12052/) (2022)
- [If it Bleeds, it Leads: A Computational Approach to Covering Crime in Los Angeles](https://mlsanity.com/paper/2206.07115/) (2022)
- [Grammatical Error Correction via Mixed-Grained Weighted Training](https://mlsanity.com/paper/2311.13848/) (2023)
- [Learning Robust Named Entity Recognizers From Noisy Data With Retrieval Augmentation](https://mlsanity.com/paper/2407.18562/) (2024)
- [JELV: A Judge of Edit-Level Validity for Evaluation and Automated Reference Expansion in Grammatical Error Correction](https://mlsanity.com/paper/2511.21700/) (2025)

## Abstract (arXiv)

Neural sequence-to-sequence (seq2seq) approaches have proven to be successful in grammatical error correction (GEC). Based on the seq2seq framework, we propose a novel fluency boost learning and inference mechanism. Fluency boosting learning generates diverse error-corrected sentence pairs during training, enabling the error correction model to learn how to improve a sentence's fluency from more instances, while fluency boosting inference allows the model to correct a sentence incrementally with multiple inference steps. Combining fluency boost learning and inference with convolutional seq2seq models, our approach achieves the state-of-the-art performance: 75.72 (F_{0.5}) on CoNLL-2014 10 annotation dataset and 62.42 (GLEU) on JFLEG test set respectively, becoming the first GEC system that reaches human-level performance (72.58 for CoNLL and 62.37 for JFLEG) on both of the benchmarks.