Talk Info: Yanan Zheng

Revisiting Few-Shot learning for Natural Language Understanding

Abstract: The few-shot learning has attracted much recent attention in the NLP community, which addresses a more practical real-world scenario when fully-supervised labels are insufficient. A key challenge lies in that prior research has been proceeding under an impractical assumption, and evaluated under a disparate set of protocols, which hinders fair comparison and measuring the progress of the field. This talk covers recent advances in few-shot learning for natural language understanding. I will first identify problems of few-shot assumptions and evaluation protocols, and then introduce and justify a practical way of few-shot evaluation. Next, by re-evaluating state-of-the-art methods on common ground, I will come to several key findings that reveal problems of the field. Finally, I will introduce several possible solutions in terms of how to improve few-shot robustness and performance.

Bio:

BioData: Yanan Zheng currently works as a postdoc in the department of computer science of Tsinghua University. In 2020, she obtained a Ph.D. degree from the School of Software, Tsinghua University. Prior to that, in 2015, she received a bachelor’s degree from Xi’an Jiaotong University. Yanan is a former research intern at Microsoft Research Asia (MSRA). She received Google China Anita Borg Scholarship. Her research interests mainly focus on natural language-based intelligence, which includes deep learning, zero-shot, and few-shot learning, natural language understanding, and generation. Additional information is available at https://zheng-yanan.github.io.