Improved Text Classification via Test-Time Augmentation

TL;DR

This paper demonstrates that carefully designed test-time augmentation policies can significantly improve the accuracy of language models in NLP tasks without additional training, by aggregating predictions over transformed test inputs.

Contribution

The authors introduce augmentation policies tailored for NLP that enhance language model performance through test-time augmentation, a technique previously underutilized in NLP.

Findings

TTA with well-designed policies improves NLP model accuracy.

Number of samples per augmentation influences TTA effectiveness.

Consistent gains over state-of-the-art methods in binary classification.

Abstract

Test-time augmentation -- the aggregation of predictions across transformed examples of test inputs -- is an established technique to improve the performance of image classification models. Importantly, TTA can be used to improve model performance post-hoc, without additional training. Although test-time augmentation (TTA) can be applied to any data modality, it has seen limited adoption in NLP due in part to the difficulty of identifying label-preserving transformations. In this paper, we present augmentation policies that yield significant accuracy improvements with language models. A key finding is that augmentation policy design -- for instance, the number of samples generated from a single, non-deterministic augmentation -- has a considerable impact on the benefit of TTA. Experiments across a binary classification task and dataset show that test-time augmentation can deliver consistent improvements over current state-of-the-art approaches.