WI-LOCNESS
收藏arXiv2025-09-30 收录
下载链接：
https://www.cl.cam.ac.uk/research/nl/bea2019st/
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是通过利用2019年BEA共享任务的系统输出构建的训练集，在移除了有偏见的输出组之后，包含了大约2500个条目的样本。此外，该数据集还用于生成训练目的的子度量分数。其规模大约为2500个条目，任务是为语法错误修正（Gec）进行模型训练。
This training dataset was constructed using the system outputs from the 2019 BEA Shared Task, and contains approximately 2500 entries after removing biased output groups. Additionally, this dataset is employed to generate sub-metric scores for training purposes, and is used for model training in the Grammatical Error Correction (GEC) task.
