Skip to yearly menu bar Skip to main content

Workshop: Efficient Natural Language and Speech Processing (Models, Training, and Inference)

Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems

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


Most prior work on task-oriented dialogue systems is restricted to supporting domain APIs. However, users may have requests that are out of the scope of these APIs. This work focuses on identifying such user requests. Existing methods for this task mainly rely on fine-tuning pre-trained models on large annotated data. We propose a novel method, REDE, based on adaptive representation learning and density estimation. REDE can be applied to zero/few-shots cases, and quickly learn a high-performing detector that is comparable to the full-supervision setting with only a few shots by updating less than 3K parameters. We demonstrate REDE's competitive performance on DSTC9 Track 1 dataset and our newly collected test set.

Chat is not available.