BriefGPT.xyz
May, 2020
少样本文本分类的动态记忆感知网络
Dynamic Memory Induction Networks for Few-Shot Text Classification
HTML
PDF
Ruiying Geng, Binhua Li, Yongbin Li, Jian Sun, Xiaodan Zhu
TL;DR
该论文提出了一种名为动态记忆诱导网络(DMIN)的模型,用于少样本文本分类。该模型利用动态路由提供更多的灵活性,以更好地适应支持集,从而提高少样本分类模型的关键能力,并在miniRCV1和ODIC数据集上实现了新的最优结果,提高了最佳性能(准确度)约2~4%。
Abstract
This paper proposes
dynamic memory induction networks
(DMIN) for few-shot
text classification
. The model utilizes dynamic routing to provide more flexibility to memory-based
→