Duolingo
收藏arXiv2025-09-30 收录
下载链接：
https://sharedtask.duolingo.com/2018.html
下载链接
链接失效反馈官方服务：
资源简介：
该数据集用于建模第二语言习得，追踪了1200名说英语的用户通过不同练习格式学习法语的过程。数据集中的稀疏特征包括用户、单词以及练习格式的独热向量。该数据集规模达到120万条记录，包含50万个独特的用户-单词对。任务是对用户在项目上的知识进行预测。
This dataset is designed for modeling second language acquisition. It tracks the French learning process of 1,200 English-speaking users across various exercise formats. The sparse features in the dataset include one-hot vectors for users, individual words, and exercise formats. The dataset has a total of 1.2 million records, encompassing 500,000 unique user-word pairs. The task is to predict users' knowledge regarding specific learning items.
提供机构：
Duolingo
