Learning How to Self-Learn: Enhancing Self-Training Using Neural Reinforcement Learning

Chen, Chenhua; Zhang, Yue; Gao, Yuze

doi:10.1109/ialp.2018.8629107

Cited by 13 publications

(6 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A somewhat similar approach is self-learning -training on examples labelled by a model itself. While it is ineffective in many settings, [3] shows that it can improve results of few-shot NER task when combined with reinforcement learning.…”

Section: Related Workmentioning

confidence: 99%

Few-shot classification in named entity recognition task

Fritzler¹,

Logacheva²,

Kretov³

2019

Proceedings of the 34th ACM/SIGAPP Symposium on Applied Computing

152

115

View full text Add to dashboard Cite

For many natural language processing (NLP) tasks the amount of annotated data is limited. This urges a need to apply semi-supervised learning techniques, such as transfer learning or meta-learning. In this work we tackle Named Entity Recognition (NER) task using Prototypical Network -a metric learning technique. It learns intermediate representations of words which cluster well into named entity classes. This property of the model allows classifying words with extremely limited number of training examples, and can potentially be used as a zero-shot learning method. By coupling this technique with transfer learning we achieve well-performing classifiers trained on only 20 instances of a target class.

show abstract

Section: Related Workmentioning

confidence: 99%

Few-shot classification in named entity recognition task

Fritzler¹,

Logacheva²,

Kretov³

2019

Proceedings of the 34th ACM/SIGAPP Symposium on Applied Computing

152

115

View full text Add to dashboard Cite

show abstract

“…Kumar et al (2010); Ma et al (2017); Li et al (2019); Mukherjee and Awadallah (2020) proposed to learn sampling weights for unlabeled data to control the selection process. Reinforcement learning (RL) methods (Chen et al, 2018;Wu et al, 2018; designed an additional Q-agent as the sample selector. Nevertheless, methods using learnable weights or RL provide marginal benefits compared to the elevated optimization cost.…”

Section: Self-trainingmentioning

confidence: 99%

Self-training Improves Pre-training for Few-shot Learning in Task-oriented Dialog Systems

Mi¹,

Zhou²,

Cai³

et al. 2021

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

As the labeling cost for different modules in task-oriented dialog (ToD) systems is expensive, a major challenge is to train different modules with the least amount of labeled data. Recently, large-scale pre-trained language models, have shown promising results for few-shot learning in ToD. In this paper, we devise a selftraining approach to utilize the abundant unlabeled dialog data to further improve state-ofthe-art pre-trained models in few-shot learning scenarios for ToD systems. Specifically, we propose a self-training approach that iteratively labels the most confident unlabeled data to train a stronger Student model. Moreover, a new text augmentation technique (GradAug) is proposed to better train the Student by replacing non-crucial tokens using a masked language model. We conduct extensive experiments and present analyses on four downstream tasks in ToD, including intent classification, dialog state tracking, dialog act prediction, and response selection. Empirical results demonstrate that the proposed self-training approach consistently improves state-of-the-art pre-trained models (BERT, ToD-BERT) when only a small number of labeled data are available.

show abstract

“…Unlike classification tasks, noisy self-labeled data can be easily eliminated by removing those which have low confidence scores; there is a lack of a comprehensive means to determine this score for a sequence-labeling data point. In several recent re-search, a deep reinforcement learning (Chen et al, 2018) and meta-learning (Wang et al, 2020) has been proposed to reduce "error propagation from noisy pseudo-labels" for sequence labeling tasks.…”

Section: Self-trainingmentioning

confidence: 99%

S-NLP at SemEval-2021 Task 5: An Analysis of Dual Networks for Sequence Tagging

Nguyên¹,

Nguyen²,

Dao³

et al. 2021

Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021)

View full text Add to dashboard Cite

The SemEval 2021 task 5: Toxic Spans Detection is a task of identifying considered-toxic spans in text, which provides a valuable, automatic tool for moderating online contents. This paper represents the second-place method for the task, an ensemble of two approaches. While one approach relies on combining different embedding methods to extract diverse semantic and syntactic representations of words in context; the other utilizes extra data with a slightly customized Self-training, a semisupervised learning technique, for sequence tagging problems. Both of our architectures take advantage of a strong language model, which was fine-tuned on a toxic classification task. Although experimental evidence indicates higher effectiveness of the first approach than the second one, combining them leads to our best results of 70.77 F1-score on the test dataset.

show abstract

Learning How to Self-Learn: Enhancing Self-Training Using Neural Reinforcement Learning

Cited by 13 publications

References 33 publications

Few-shot classification in named entity recognition task

Few-shot classification in named entity recognition task

Self-training Improves Pre-training for Few-shot Learning in Task-oriented Dialog Systems

S-NLP at SemEval-2021 Task 5: An Analysis of Dual Networks for Sequence Tagging

Contact Info

Product

Resources

About