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


Abstract
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-shot cases, and quickly learns a high-performing detector with only a few shots by updating less than 3K parameters. We demonstrate REDE’s competitive performance on DSTC9 data and our newly collected test set.
Anthology ID:
2021.nlp4convai-1.27
Volume:
Proceedings of the 3rd Workshop on Natural Language Processing for Conversational AI
Month:
November
Year:
2021
Address:
Online
Venues:
EMNLP | NLP4ConvAI
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
281–288
Language:
URL:
https://aclanthology.org/2021.nlp4convai-1.27
DOI:
10.18653/v1/2021.nlp4convai-1.27
Bibkey:
Cite (ACL):
Di Jin, Shuyang Gao, Seokhwan Kim, Yang Liu, and Dilek Hakkani-Tur. 2021. Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems. In Proceedings of the 3rd Workshop on Natural Language Processing for Conversational AI, pages 281–288, Online. Association for Computational Linguistics.
Cite (Informal):
Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems (Jin et al., NLP4ConvAI 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.nlp4convai-1.27.pdf
Code
 jind11/rede