Dynamic Prototype Selection by Fusing Attention Mechanism for Few-Shot Relation Classification

Linfang Wu, Hua Ping Zhang*, Yaofei Yang, Xin Liu, Kai Gao

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

7 引用 (Scopus)

摘要

In a relation classification task, few-shot learning is an effective method when the number of training instances decreases. The prototypical network is a few-shot classification model that generates a point to represent each class, and this point is called a prototype. The mean is used to select prototypes for each class from a support set in a prototypical network. This method is fixed and static, and will lose some information at the sentence level. Therefore, we treat the mean selection as a special attention mechanism, then we expand the mean selection to dynamic prototype selection by fusing a self-attention mechanism. We also propose a query-attention mechanism to more accurately select prototypes. Experimental results on the FewRel dataset show that our model achieves significant and consistent improvements to baselines on few-shot relation classification.

源语言英语
主期刊名Intelligent Information and Database Systems - 12th Asian Conference, ACIIDS 2020, Proceedings
编辑Ngoc Thanh Nguyen, Bogdan Trawinski, Kietikul Jearanaitanakij, Suphamit Chittayasothorn, Ali Selamat
出版商Springer
431-441
页数11
ISBN(印刷版)9783030419639
DOI
出版状态已出版 - 2020
活动12th Asian Conference on Intelligent Information and Database Systems, ACIIDS 2020 - Phuket, 泰国
期限: 23 3月 202026 3月 2020

出版系列

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
12033 LNAI
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议12th Asian Conference on Intelligent Information and Database Systems, ACIIDS 2020
国家/地区泰国
Phuket
时期23/03/2026/03/20

指纹

探究 'Dynamic Prototype Selection by Fusing Attention Mechanism for Few-Shot Relation Classification' 的科研主题。它们共同构成独一无二的指纹。

引用此