A Neural Few-Shot Text Classification Reality Check

TL;DR

This paper evaluates various neural few-shot text classification models, including adaptations from computer vision, revealing that recent models often underperform older methods on complex NLP tasks despite using transformers.

Contribution

It compares adapted computer vision models and transformer-based models on NLP tasks, providing a comprehensive evaluation framework and revealing surprising performance insights.

Findings

Recent models perform worse than older ones on intent detection.

Transformers alone do not guarantee superior performance in few-shot NLP.

A simple baseline model is surprisingly effective.

Abstract

Modern classification models tend to struggle when the amount of annotated data is scarce. To overcome this issue, several neural few-shot classification models have emerged, yielding significant progress over time, both in Computer Vision and Natural Language Processing. In the latter, such models used to rely on fixed word embeddings before the advent of transformers. Additionally, some models used in Computer Vision are yet to be tested in NLP applications. In this paper, we compare all these models, first adapting those made in the field of image processing to NLP, and second providing them access to transformers. We then test these models equipped with the same transformer-based encoder on the intent detection task, known for having a large number of classes. Our results reveal that while methods perform almost equally on the ARSC dataset, this is not the case for the Intent Detection task, where the most recent and supposedly best competitors perform worse than older and simpler ones (while all are given access to transformers). We also show that a simple baseline is surprisingly strong. All the new developed models, as well as the evaluation framework, are made publicly available.