MixText: Linguistically-Informed Interpolation of Hidden Space for Semi-Supervised Text Classification

TL;DR

MixText introduces a semi-supervised text classification method that leverages hidden space interpolation and data augmentation to improve performance, especially with limited labeled data.

Contribution

The paper proposes TMix, a novel data augmentation technique in hidden space, and demonstrates how mixing labeled, unlabeled, and augmented data enhances semi-supervised learning.

Findings

Outperforms state-of-the-art semi-supervised methods on benchmarks.

Significant gains when supervision is extremely limited.

Publicly available code for reproducibility.

Abstract

This paper presents MixText, a semi-supervised learning method for text classification, which uses our newly designed data augmentation method called TMix. TMix creates a large amount of augmented training samples by interpolating text in hidden space. Moreover, we leverage recent advances in data augmentation to guess low-entropy labels for unlabeled data, hence making them as easy to use as labeled data.By mixing labeled, unlabeled and augmented data, MixText significantly outperformed current pre-trained and fined-tuned models and other state-of-the-art semi-supervised learning methods on several text classification benchmarks. The improvement is especially prominent when supervision is extremely limited. We have publicly released our code at https://github.com/GT-SALT/MixText.