How to choose "Good" Samples for Text Data Augmentation

TL;DR

This paper introduces a self-training selection framework with two selectors to identify high-quality augmented samples for text classification, improving data augmentation effectiveness.

Contribution

It proposes a novel sample selection method combining entropy-based filtering with recall strategies based on word overlap and semantic similarity.

Findings

Effective in selecting high-quality augmented samples

Improves text classification performance with data augmentation

Simple and efficient framework

Abstract

Deep learning-based text classification models need abundant labeled data to obtain competitive performance. Unfortunately, annotating large-size corpus is time-consuming and laborious. To tackle this, multiple researches try to use data augmentation to expand the corpus size. However, data augmentation may potentially produce some noisy augmented samples. There are currently no works exploring sample selection for augmented samples in nature language processing field. In this paper, we propose a novel self-training selection framework with two selectors to select the high-quality samples from data augmentation. Specifically, we firstly use an entropy-based strategy and the model prediction to select augmented samples. Considering some samples with high quality at the above step may be wrongly filtered, we propose to recall them from two perspectives of word overlap and semantic similarity. Experimental results show the effectiveness and simplicity of our framework.