Meta-Learning Adversarial Domain Adaptation Network for Few-Shot Text Classification

Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 1664–1673 August 1–6, 2021. ©2021 Association for Computational Linguistics 1664 Meta-Learning Adversarial Domain Adaptation Network for Few-Shot Text Classification ChengCheng Han1 Zeqiu Fan1 Dongxiang Zhang2 Minghui Qiu3 Ming Gao1∗ Aoying Zhou1 1School of Data Science and Engineering, East China Normal University 2College of Computer Science and Technology, Zhejiang University 3Alibaba Group {51195100009,51195100007}@stu.ecnu.edu.cn zhangdongxiang@zju.edu.cn minghui.qmh@alibaba-inc.com mgao@dase.ecnu.edu.cn ayzhou@sei.ecnu.edu.cn Abstract Meta-learning has emerged as a trending tech- nique to tackle few-shot text classification and achieved state-of-the-art performance. How- ever, existing solutions heavily rely on the ex- ploitation of lexical features and their distri- butional signatures on training data, while ne- glecting to strengthen the model’s ability to adapt to new tasks. In this paper, we pro- pose a novel meta-learning framework inte- grated with an adversarial domain adaptation network, aiming to improve the adaptive abil- ity of the model and generate high-quality text embedding for new classes. Extensive ex- periments are conducted on four benchmark datasets and our method demonstrates clear su- periority over the state-of-the-art models in all the datasets. In particular, the accuracy of 1- shot and 5-shot classification on the dataset of 20 Newsgroups is boosted from 52.1% to 59.6%, and from 68.3% to 77.8%, respec- tively1. 1 Introduction Few-shot text classification (Yu et al., 2018; Geng et al., 2019) is a task in which a model will be adapted to predict new classes not seen in training. For each of these new classes, we only have a few labeled examples. To be specific, we are given lots of training data with a set of classes Ytrain. After training, our goal is to get accurate classification results on the testing data with a set of new classes Ytest, which is disjoint to Ytrain. Only a small la- beled support set will be available in the testing ∗Corresponding author 1The source code of the paper is available at https:// github.com/hccngu/MLADA. stage. If the support set contains K labeled exam- ples for each of the N unique classes, we refer to the task as a N-way K-shot classification. Existing approaches for few-shot text classifica- tion mainly fall into two categories: (1) transfer- learning based methods (Howard and Ruder, 2018; Pan et al., 2019; Gupta et al., 2020), which aim to transfer knowledge learned from a task to a new task or leverage general-domain pretraining and fine-tuning techniques for few-shot classifi- cation. (2) meta-learning based methods (Jamal et al., 2018; Yu et al., 2018; Geng et al., 2019, 2020; Bao et al., 2020), which aim to learn generic information (meta-knowledge) by recreating train- ing episodes, so that it can classify new classes through only a few labeled examples. Among these methods, Bao et al. (2020) leveraged distributional signatures (e.g. word frequency and information entropy) to train a model within a meta-learning framework, and achieved state-of-the-art perfor- mance. However, the method pays more atten- tion to statistical information and ignores other implicit information such as correlation between words. Furthermore, existing meta-learning meth- ods heavily …

Content truncated to 500 words