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

Han, Chengcheng; Fan, Zeqiu; Zhang, Dongxiang; Qiu, Minghui; Gao, Ming; Zhou, Aoying

doi:10.18653/v1/2021.findings-acl.145

Cited by 27 publications

(21 citation statements)

References 17 publications

(29 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MLADA [4] is committed to improving the model's adaptability. The text representations are obtained by introducing an adversarial domain adaptation network.…”

Section: Classification Experiments Resultsmentioning

confidence: 99%

“…In recent years, few-shot learning has been generously applied to computer vision tasks and is emerging as a promising solution to the low-resource regime [2]. Some researches have achieved better results on text classification tasks [3,4] by focusing on the local-to-global paradigm, which minimizes the distance between the support and query distributions to obtain strong performance over the entire dataset. Moreover, feature selection is a way to reduce time consumption.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Few-Shot Text Classification with Global–Local Feature Information

Wang

Cheng

et al. 2022

Sensors

View full text Add to dashboard Cite

Meta-learning frameworks have been proposed to generalize machine learning models for domain adaptation without sufficient label data in computer vision. However, text classification with meta-learning is less investigated. In this paper, we propose SumFS to find global top-ranked sentences by extractive summary and improve the local vocabulary category features. The SumFS consists of three modules: (1) an unsupervised text summarizer that removes redundant information; (2) a weighting generator that associates feature words with attention scores to weight the lexical representations of words; (3) a regular meta-learning framework that trains with limited labeled data using a ridge regression classifier. In addition, a marine news dataset was established with limited label data. The performance of the algorithm was tested on THUCnews, Fudan, and marine news datasets. Experiments show that the SumFS can maintain or even improve accuracy while reducing input features. Moreover, the training time of each epoch is reduced by more than 50%.

show abstract

“…MLADA [4] is committed to improving the model's adaptability. The text representations are obtained by introducing an adversarial domain adaptation network.…”

Section: Classification Experiments Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Few-Shot Text Classification with Global–Local Feature Information

Wang

Cheng

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…We evaluate our models on typical 5-way 1-shot and 5way 5-shot text classification settings. Following the evaluation setup in (Dopierre, Gravier, and Logerais 2021), we report the average accuracy over 600 episodes sampled from the test set for intent classification datasets; and following (Han et al 2021), we report the average accuracy over 1000 episodes sampled from the test set for news or review classification datasets. We run each experimental setting 5 times.…”

Section: Experimental Settingsmentioning

confidence: 99%

“…Meta-learning has been extensively studied in image classification and achieve remarkable successes (Vinyals et al 2016;Snell, Swersky, and Zemel 2017;Finn, Abbeel, and Levine 2017;Sung et al 2018;Hou et al 2019;Tseng et al 2020;Liu et al 2021;Gao et al 2021). The effectiveness in image classification motivates the recent application of meta-learning to few-shot text classification (Yu et al 2018;Geng et al 2019Geng et al , 2020Bao et al 2020;Han et al 2021).…”

Section: Introductionmentioning

confidence: 99%

ContrastNet: A Contrastive Learning Framework for Few-Shot Text Classification

Chen

Zhang

Mao

et al. 2022

AAAI

View full text Add to dashboard Cite

Few-shot text classification has recently been promoted by the meta-learning paradigm which aims to identify target classes with knowledge transferred from source classes with sets of small tasks named episodes. Despite their success, existing works building their meta-learner based on Prototypical Networks are unsatisfactory in learning discriminative text representations between similar classes, which may lead to contradictions during label prediction. In addition, the task-level and instance-level overfitting problems in few-shot text classification caused by a few training examples are not sufficiently tackled. In this work, we propose a contrastive learning framework named ContrastNet to tackle both discriminative representation and overfitting problems in few-shot text classification. ContrastNet learns to pull closer text representations belonging to the same class and push away text representations belonging to different classes, while simultaneously introducing unsupervised contrastive regularization at both task-level and instance-level to prevent overfitting. Experiments on 8 few-shot text classification datasets show that ContrastNet outperforms the current state-of-the-art models.

show abstract

“…It has been receiving increasing attention for its potential in reducing data collection effort and computational costs and extending to rare cases. To deal with data-scarcity in NLU problems, previous research introduces external knowledge (Sui et al, 2021), utilizes meta-learning (Geng et al, 2019;Han et al, 2021) and adopts data augmentation to generate labeled utterances for few-shot classes (Murty et al, 2021;Wei et al, 2021). Recent studies (Radford et al, 2019;Brown et al, 2020) have shown that large-scale pre-trained language models are able to perform NLU tasks in a few-shot learning manner.…”

Section: Related Workmentioning

confidence: 99%

Imagination-Augmented Natural Language Understanding

Lu¹,

Zhu²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Human brains integrate linguistic and perceptual information simultaneously to understand natural language, and hold the critical ability to render imaginations. Such abilities enable us to construct new abstract concepts or concrete objects, and are essential in involving practical knowledge to solve problems in low-resource scenarios. However, most existing methods for Natural Language Understanding (NLU) are mainly focused on textual signals. They do not simulate human visual imagination ability, which hinders models from inferring and learning efficiently from limited data samples. Therefore, we introduce an Imagination-Augmented Cross-modal Encoder (iACE) to solve natural language understanding tasks from a novel learning perspective-imagination-augmented cross-modal understanding. iACE enables visual imagination with external knowledge transferred from the powerful generative and pre-trained vision-and-language models. Extensive experiments on GLUE SWAG (Zellers et al., 2018) show that iACE achieves consistent improvement over visually-supervised pre-trained models. More importantly, results in extreme and normal few-shot settings validate the effectiveness of iACE in low-resource natural language understanding circumstances. 1

show abstract

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

Cited by 27 publications

References 17 publications

Few-Shot Text Classification with Global–Local Feature Information

Few-Shot Text Classification with Global–Local Feature Information

ContrastNet: A Contrastive Learning Framework for Few-Shot Text Classification

Imagination-Augmented Natural Language Understanding

Contact Info

Product

Resources

About