Towards Textual Out-of-Domain Detection without In-Domain Labels

TL;DR

This paper introduces a novel unsupervised method for out-of-domain detection in text, which outperforms likelihood-based approaches and rivals supervised methods without requiring in-domain labels.

Contribution

It proposes a new representation learning approach combining clustering and contrastive learning for OOD detection without in-domain labels.

Findings

The proposed method significantly outperforms likelihood-based approaches.

It is competitive with state-of-the-art supervised methods.

Extensive experiments validate the effectiveness of the approach.

Abstract

In many real-world settings, machine learning models need to identify user inputs that are out-of-domain (OOD) so as to avoid performing wrong actions. This work focuses on a challenging case of OOD detection, where no labels for in-domain data are accessible (e.g., no intent labels for the intent classification task). To this end, we first evaluate different language model based approaches that predict likelihood for a sequence of tokens. Furthermore, we propose a novel representation learning based method by combining unsupervised clustering and contrastive learning so that better data representations for OOD detection can be learned. Through extensive experiments, we demonstrate that this method can significantly outperform likelihood-based methods and can be even competitive to the state-of-the-art supervised approaches with label information.