A Formal Framework for Fluency-based Multi-Reference Evaluation in Grammatical Error Correction
This addresses the need for more flexible and multilingual evaluation metrics in grammatical error correction, though it is incremental as it builds on existing frameworks like GLEU.
The paper tackled the problem of evaluating grammatical error correction by proposing a formal framework for fluency-based multi-reference evaluation, which aggregates n-gram similarity over multiple valid corrections, and showed empirical results on Czech, Estonian, Ukrainian, and Chinese corpora that the strategies capture complementary aspects of fluency and coverage.
Evaluating grammatical error correction requires metrics that reflect the diversity of valid human corrections rather than privileging a single reference. Existing frameworks, largely edit-based and English-centric, rely on rigid alignments between system and reference edits, limiting their applicability in multilingual and generative settings. This paper introduces a formal framework for \textit{fluency-based multi-reference evaluation}, framing $n$-gram similarity as an aggregation problem over multiple legitimate corrections. Within this formulation, we instantiate GLEU through four aggregation strategies--\textsc{select-best}, \textsc{simple-average}, \textsc{weighted-average}, and \textsc{merged-counts}--and analyze their properties of boundedness, monotonicity, and sensitivity to reference variation. Empirical results on Czech, Estonian, Ukrainian, and Chinese corpora show that these strategies capture complementary aspects of fluency and coverage. The framework unifies multi-reference evaluation into a principled, fluency-oriented approach that incorporates linguistic diversity without penalizing legitimate variation.