arXiv cs.CL· Kevin Fan, Eric Yun·· 12 小时前AI 评分31
CL-MEP：用匹配作文对对比学习缓解 ESL 学习者自动作文评分偏差
Mitigating Bias in Automated Essay Scoring for ESL Learners via Contrastive Learning
AI 导读
研究提出匹配作文对对比学习（CL-MEP）双向对齐策略，将自动作文评分系统对高熟练度 ESL 学习者的评分偏差降低 39.9%，同时提升整体准确率。该方法成功将合理的句法复杂度与表层语法错误区分开来。
正文 · 原文
Abstract:Automated Essay Scoring systems disproportionately penalize high-proficiency English as a Second Language (ESL) learners. We propose Contrastive Learning with Matched Essay Pairs (CL-MEP), a bi-directional alignment strategy. CL-MEP reduces this scoring bias by 39.9% while improving overall accuracy, successfully disentangling valid syntactic complexity from surface-level grammatical errors.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2601.16724 [cs.CL] |
| (or arXiv:2601.16724v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2601.16724 arXiv-issued DOI via DataCite |
Submission history
From: Kevin Fan [view email]
[v1]
Fri, 23 Jan 2026 13:18:30 UTC (10 KB)
[v2]
Mon, 5 Oct 2026 05:06:57 UTC (13 KB)
[v3]
Sun, 1 Nov 2026 21:12:41 UTC (13 KB)
来源：arXiv cs.CL · arxiv.org