WBP3l
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/muhaochen/bilingual_dictionaries
下载链接
链接失效反馈官方服务：
资源简介：
该数据集名为WBP3l，包含了6000对双语句子，旨在用于英法及英西跨语言句子同义辨识任务。这些句子对是通过跨语言链接对齐的描述，排除了Wikt3l中的单词描述，以用于训练。该数据集的规模为6000对，所涉及的任务是双语同义辨识。
The dataset named WBP3l consists of 6000 bilingual sentence pairs, and is intended for cross-lingual sentence paraphrase identification tasks between English and French, as well as English and Spanish. These sentence pairs are aligned descriptive texts obtained via cross-lingual links, with the word-level descriptions from Wikt3l excluded for training use. With 6000 pairs in total, this dataset targets the bilingual paraphrase identification task.
