Few-Shot Text Classification with Triplet Networks, Data Augmentation, and Curriculum Learning

TL;DR

This paper enhances few-shot text classification by combining triplet networks with data augmentation and curriculum learning, leading to faster training and improved accuracy on multiple tasks.

Contribution

It introduces curriculum data augmentation, a novel training strategy that effectively integrates augmented data to boost few-shot text classification performance.

Findings

Data augmentation improves triplet network accuracy by up to 3%.

Curriculum data augmentation accelerates training and enhances robustness.

Two-stage and gradual schedules outperform single-stage training.

Abstract

Few-shot text classification is a fundamental NLP task in which a model aims to classify text into a large number of categories, given only a few training examples per category. This paper explores data augmentation -- a technique particularly suitable for training with limited data -- for this few-shot, highly-multiclass text classification setting. On four diverse text classification tasks, we find that common data augmentation techniques can improve the performance of triplet networks by up to 3.0% on average. To further boost performance, we present a simple training strategy called curriculum data augmentation, which leverages curriculum learning by first training on only original examples and then introducing augmented data as training progresses. We explore a two-stage and a gradual schedule, and find that, compared with standard single-stage training, curriculum data augmentation trains faster, improves performance, and remains robust to high amounts of noising from augmentation.