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

Di Jin, Shuyang Gao, Seokhwan Kim, Yang Liu, Dilek Hakkani-Tur

Research output: Contribution to journalArticlepeer-review


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.

Original languageEnglish (US)
Pages (from-to)1386-1395
Number of pages10
JournalIEEE/ACM Transactions on Audio Speech and Language Processing
StatePublished - 2022
Externally publishedYes


  • natural language processing
  • Out-of-domain detection
  • unsupervised representation learning

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Acoustics and Ultrasonics
  • Computational Mathematics
  • Electrical and Electronic Engineering


Dive into the research topics of 'Towards Textual Out-of-Domain Detection Without In-Domain Labels'. Together they form a unique fingerprint.

Cite this