Grammatical Error Correction
Overview
Grammatical error correction studies how to identify and fix errors in written or spoken language while preserving the writer’s intended meaning. Common approaches include neural sequence-to-sequence models trained on parallel or synthetically corrupted text, as well as large language models guided by prompts, retrieval, fine-tuning, or reinforcement learning. Systems are tested on benchmarks such as CoNLL-2014 and BEA-2019, with growing work on multilingual, low-resource, spoken, and domain-specific settings. A central challenge is balancing precision and minimal edits against fluency: systems must avoid over-correction, handle valid alternatives that differ from a reference, and provide reliable evaluation and useful feedback.
Latest papers
23 papers1 paper in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.
Precision--Recall Trade-off on the Inference Level for LLMs in Grammatical Error Correction
Beyond Reference-Based Evaluation: Reward Models for Meta-Evaluation of Grammatical Error Correction
Larger Context Window, Fewer Overcorrections: Optimizing Prompts and Batching for Minimal-Edit Grammatical Error Correction
Contextual Tamil Spelling and Grammar Correction Using Progressively Fine-Tuned Sequence-to-Sequence Transformers
A Layered Taxonomy for Chinese Learner Grammatical Error Annotation
Clause Encounters of the Third Kind: Can LLMs Replace Language Teachers?
Accurate but Natural? Diagnosing Grammatical and Idiomatic Gaps in Japanese EFL Writing
Encode Errors: Representational Retrieval of In-Context Demonstrations for Multilingual Grammatical Error Correction
CoCoGEC: Counterfactual Generation for Robust Grammatical Error Correction
ArabiGEE: A Hierarchical Taxonomy for Arabic Grammatical Error Explanation
How Far Can Prompting Go for Minimal-Edit Ukrainian Grammatical Error Correction?
Refining Word-Based Grammatical Error Annotation for L2 Korean
m2 edits. We then define a Korean ERRANT-style annotation scheme that preserves the MRU core while distinguishing functional morpheme errors, spelling errors, word boundary errors, and word order errors. We also augment the KoLLA corpus with an additional reference correction, yielding a multi-reference evaluation setting for Korean GEC. Empirical validation shows that the refined NIKL targets yield lower perplexity, the converted m2 files achieve higher agreement with source-target edit representations, and the refined resources improve KoBART-based correction under the same model setting. Multi-reference KoLLA evaluation further reduces the penalty imposed on valid corrections that diverge from a single reference, especially for neural and prompted GEC systems. These results show that Korean GEC evaluation depends not only on correction models, but also on reference data and edit annotations that reflect Korean morphology, spacing, and correction variability.