EPIE Dataset
收藏arXiv2020-06-17 更新2024-06-21 收录
下载链接：
https://github.com/prateeksaxena2809/EPIE
下载链接
链接失效反馈官方服务：
资源简介：
EPIE数据集是由海得拉巴国际信息技术研究所创建的，包含25206条标注了717个习语表达的句子。该数据集通过自动系统高精度标注，旨在解决自然语言处理中习语理解的难题。数据集分为静态和正式两种习语类型，分别处理不同类型的习语表达。该数据集的应用领域包括机器翻译和习语检测，旨在提高这些任务中习语处理的准确性和效率。
The EPIE dataset was created by the International Institute of Information Technology, Hyderabad, and consists of 25,206 sentences annotated with 717 idiomatic expressions. Annotated with high precision via automated systems, this dataset is designed to address the challenges of idiom understanding in natural language processing (NLP). It is categorized into two idiom types: static and formal, which respectively handle different categories of idiomatic expressions. Its application fields include machine translation and idiom detection, aiming to improve the accuracy and efficiency of idiom processing in these tasks.
提供机构：
海得拉巴国际信息技术研究所
创建时间：
2020-06-17
