Mask-guided BERT for few-shot text classification

Wenxiong Liao, Zhengliang Liu, Haixing Dai, Zihao Wu, Yiyang Zhang, Xiaoke Huang, Yuzhong Chen, Xi Jiang, David Liu, Dajiang Zhu, Sheng Li, Wei Liu, Tianming Liu, Quanzheng Li, Hongmin Cai, Xiang Li

Abstract: Highlights•A simple and modular framework, ”Mask-BERT”, for few-shot text classification.•Mask-BERT itilizes selective masking of text inputs to filter out irrelevant information.•Mask-BERT guides the model towards relevant information to enhance few-shot learning.