Sequence Level Mixed Sample Data Augmentation

EMNLP 2020, pp. 5547-5552, 2020.

被引用0|引用|浏览39|DOI:https://doi.org/10.18653/V1/2020.EMNLP-MAIN.447
其它链接arxiv.org|academic.microsoft.com

摘要

Despite their empirical success, neural networks still have difficulty capturing compositional aspects of natural language. This work proposes a simple data augmentation approach to encourage compositional behavior in neural models for sequence-to-sequence problems. Our approach, SeqMix, creates new synthetic examples by softly combining ...更多

代码

数据

下载 PDF 全文
您的评分 :
0

 

标签
评论