QShane/CL2GEC
收藏Hugging Face2025-11-12 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/QShane/CL2GEC
下载链接
链接失效反馈官方服务：
资源简介：
CL²GEC是一个面向中文学术写作语法错误修正的连续学习协议基准，包含10个一级学科的语料，每个样本包括一个错误句子和至少一个修正后的参考句子。数据集旨在用于研究学术文本中的语法错误修正、跨领域建模的鲁棒性、以及学科感知模型，并专注于连续学习中的遗忘和迁移问题。
CL²GEC is a benchmark for continual learning in Chinese grammatical error correction for scholarly writing, covering 10 first-level disciplines. Each sample includes an errorful sentence and at least one corrected reference sentence. The dataset is intended for research on grammatical error correction in academic texts, robustness of cross-domain modeling, discipline-aware modeling, and focusing on issues of forgetting and transfer in continual learning.
提供机构：
QShane
