遇见数据集

Turku Paraphrase Corpus

收藏
DataCite Commons2022-06-01 更新2024-07-13 收录
官方服务:

资源简介:

The Turku Paraphrase corpus consists of over 100,000+ manually selected Finnish paraphrases, most of which are in their document context. The paraphrases are manually classified using a scheme capturing the degree of contextual dependence, as well as a possible subsumption relation and other flags such as style and minor syntactic differences. The corpus also contains a small Swedish test set.

提供机构:
ELG
创建时间:
2022-06-01
二维码
社区交流群
二维码
科研交流群
商业服务