Few-shot learning through contextual data augmentation

2021-03-31 · EACL 2021 2 · Farid Arthaud, Rachel Bawden, Alexandra Birch

Machine translation (MT) models used in industries with constantly changing topics, such as translation or news agencies, need to adapt to new data to maintain their performance over time. Our aim is to teach a pre-trained MT model to translate previously unseen words accurately, based on very few examples. We propose (i) an experimental setup allowing us to simulate novel vocabulary appearing in human-submitted translations, and (ii) corresponding evaluation metrics to compare our approaches. We extend a data augmentation approach using a pre-trained language model to create training examples with similar contexts for novel words. We compare different fine-tuning and data augmentation approaches and show that adaptation on the scale of one to five examples is possible. Combining data augmentation with randomly selected training sentences leads to the highest BLEU score and accuracy improvements. Impressively, with only 1 to 5 examples, our model reports better accuracy scores than a reference system trained with on average 313 parallel examples.

📄 PDF Abstract BibTeX arXiv:2103.16911

Code (1)

https://gitlab.com/farid-fari/fewshot-learning 공식 구현

Tasks

Data Augmentation Few-Shot LearningLanguage ModelingLanguage Modelling Machine Translation Translation

Contextual Representation Anchor Network to Alleviate Selection Bias in Few-Shot Drug Discovery

2024-10-28 · Ruifeng Li, Wei Liu, Xiangxin Zhou, Mingqian Li 외

In the drug discovery process, the low success rate of drug candidate screening often leads to insufficient labeled data, causing the few-shot learning problem in molecular property prediction. Existing methods for few-s…

Drug Discovery Few-Shot Learning Molecular Property Prediction Property Prediction+1

LDCA: Local Descriptors with Contextual Augmentation for Few-Shot Learning

2024-01-24 · Maofa Wang, Bingchen Yan

Few-shot image classification has emerged as a key challenge in the field of computer vision, highlighting the capability to rapidly adapt to new tasks with minimal labeled data. Existing methods predominantly rely on im…

Classification Few-Shot Image Classification Few-Shot Learning image-classification+1

CoSDA-ML: Multi-Lingual Code-Switching Data Augmentation for Zero-Shot Cross-Lingual NLP

2020-06-11 · Libo Qin, Minheng Ni, Yue Zhang, Wanxiang Che

Multi-lingual contextualized embeddings, such as multilingual-BERT (mBERT), have shown success in a variety of zero-shot cross-lingual tasks. However, these models are limited by having inconsistent contextualized repres…

Data Augmentation

LLM-DA: Data Augmentation via Large Language Models for Few-Shot Named Entity Recognition

2024-02-22 · Junjie Ye, Nuo Xu, Yikun Wang, Jie zhou 외

Despite the impressive capabilities of large language models (LLMs), their performance on information extraction tasks is still not entirely satisfactory. However, their remarkable rewriting capabilities and extensive wo…

Data Augmentation few-shot-ner Few-shot NER named-entity-recognition+3

ProtoConNet: Prototypical Augmentation and Alignment for Open-Set Few-Shot Image Classification

2025-07-16 · Kexuan Shi, Zhuang Qi, Jingjing Zhu, Lei Meng 외 arxiv

Open-set few-shot image classification aims to train models using a small amount of labeled data, enabling them to achieve good generalization when confronted with unknown environments. Existing methods mainly use visual…

Few-Shot Image Classification Representation Learning