Zero-Shot Text Classification via Self-Supervised Tuning

TL;DR

This paper introduces a self-supervised tuning method for language models that improves zero-shot text classification by leveraging unlabeled data and a novel first sentence prediction task, outperforming existing approaches on multiple benchmarks.

Contribution

The paper proposes a new self-supervised learning paradigm for zero-shot classification that reduces prompt sensitivity and eliminates the need for annotated task data.

Findings

Outperforms state-of-the-art on 7 out of 10 tasks

Less sensitive to prompt design

Effective with unlabeled data

Abstract

Existing solutions to zero-shot text classification either conduct prompting with pre-trained language models, which is sensitive to the choices of templates, or rely on large-scale annotated data of relevant tasks for meta-tuning. In this work, we propose a new paradigm based on self-supervised learning to solve zero-shot text classification tasks by tuning the language models with unlabeled data, called self-supervised tuning. By exploring the inherent structure of free texts, we propose a new learning objective called first sentence prediction to bridge the gap between unlabeled data and text classification tasks. After tuning the model to learn to predict the first sentence in a paragraph based on the rest, the model is able to conduct zero-shot inference on unseen tasks such as topic classification and sentiment analysis. Experimental results show that our model outperforms the state-of-the-art baselines on 7 out of 10 tasks. Moreover, the analysis reveals that our model is less sensitive to the prompt design. Our code and pre-trained models are publicly available at https://github.com/DAMO-NLP-SG/SSTuning .