arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2108.06332cs.CL
收藏

FlipDA:用于少样本学习的有效且稳健的数据增强

FlipDA: Effective and Robust Data Augmentation for Few-Shot Learning

Tsinghua University(清华大学);Shanghai Qi Zhi Institute(上海期智研究院);Beijing Academy of Artificial Intelligence(北京智源人工智能研究院)

Jing Zhou, Yanan Zheng, Jie Tang, Jian Li, Zhilin Yang

摘要:

大多数以前的文本数据增强方法仅限于简单的任务和弱基线。我们在硬任务上探索数据增强(即,Few-Shot自然语言理解)和强基线(即,具有超过十亿个参数的预训练模型)。在此设置下,我们复制了大量以前的增强方法,发现这些方法带来的边际收益最好,有时会降低性能很多。为了解决这一挑战,我们提出了一种新的数据增强方法FlipDA,它联合使用生成模型和分类器来生成标签翻转数据。FlipDA的核心思想是发现生成标签翻转数据比生成标签保留数据对性能更重要。实验表明,FlipDA在有效性和鲁棒性之间实现了良好的平衡-它大大改善了许多任务,同时不会对其他任务产生负面影响。

英文摘要:

Most previous methods for text data augmentation are limited to simple tasks and weak baselines. We explore data augmentation on hard tasks (i.e., few-shot natural language understanding) and strong baselines (i.e., pretrained models with over one billion parameters). Under this setting, we reproduced a large number of previous augmentation methods and found that these methods bring marginal gains at best and sometimes degrade the performance much. To address this challenge, we propose a novel data augmentation method FlipDA that jointly uses a generative model and a classifier to generate label-flipped data. Central to the idea of FlipDA is the discovery that generating label-flipped data is more crucial to the performance than generating label-preserved data. Experiments show that FlipDA achieves a good tradeoff between effectiveness and robustness -- it substantially improves many tasks while not negatively affecting the others.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑