wi_locness

Cambridge English Write & Improve + LOCNESS 提供英文文本数据，规模在1K到10K之间，包含母语和非母语英语学习者的作文。数据集中包含文本的CEFR等级信息，并由专家生成了语法错误标注，标注信息包括错误的起始位置、结束位置以及修改后的文本。此数据集主要用于语法纠错任务，旨在帮助语言学习者提高写作技能，也可用于提醒母语使用者注意意外错误。数据集遵循特定的许可协议，包括Write & Improve License和LOCNESS License，限制了商业用途和数据再分发。Cambridge English Write & Improve and LOCNESS offer English textual data, with the dataset size ranging from 1,000 to 10,000 instances. The dataset contains essays written by both native and non-native English learners, and includes CEFR level information for each text. It features grammatical error annotations generated by experts, which cover the start and end positions of errors as well as the corrected text. This dataset is primarily used for grammatical error correction tasks, aiming to help language learners improve their writing skills, and can also be used to remind native English speakers of accidental grammatical errors. The dataset is subject to specific license agreements including the Write & Improve License and the LOCNESS License, which restrict commercial use and data redistribution.

搜集汇总

背景与挑战

背景概述

wi_locness数据集是一个包含1K到10K规模英文文本的数据集，涵盖母语和非母语英语学习者的作文，附带CEFR等级和专家标注的语法错误信息，主要用于语法纠错任务，遵循非商业用途的许可协议。