ntunlp / daga

Data Augmentation with a Generation Approach for Low-resource Tagging Tasks
MIT License
79 stars 15 forks source link

实验数据来源 #6

Open GaryYufei opened 3 years ago

GaryYufei commented 3 years ago

您好!感谢您这个工作启发了我最近的一些思路。我打算复现一下这个实验,但我发现您这边CONLL03上面的full-data效果只有83左右的F1,即便是在2003年,CONLL03上面F1就已经能够到88左右。我想跟您确认一下您用的是哪个数据呢?您能给个NER实验数据的链接吗?

感谢您的帮助!

Bosheng2020 commented 3 years ago

We used the standard CoNLL03 English dataset ( https://www.clips.uantwerpen.be/conll2003/ner/ ). You may try our NER tagging model posted on https://github.com/ntunlp/daga