LAD: Language Models as Data for Zero-Shot Dialog

Shikib Mehri, Yasemin Altun, Maxine Eskenazi


Abstract
To facilitate zero-shot generalization in task-oriented dialog, this paper proposes Language Models as Data (LAD). LAD is a paradigm for creating diverse and accurate synthetic data which conveys the necessary structural constraints and can be used to train a downstream neural dialog model. LAD leverages GPT-3 to induce linguistic diversity. LAD achieves significant performance gains in zero-shot settings on intent prediction (+15%), slot filling (+31.4 F-1) and next action prediction (+10 F-1). Furthermore, an interactive human evaluation shows that training with LAD is competitive with training on human dialogs.
Anthology ID:
2022.sigdial-1.55
Volume:
Proceedings of the 23rd Annual Meeting of the Special Interest Group on Discourse and Dialogue
Month:
September
Year:
2022
Address:
Edinburgh, UK
Editors:
Oliver Lemon, Dilek Hakkani-Tur, Junyi Jessy Li, Arash Ashrafzadeh, Daniel Hernández Garcia, Malihe Alikhani, David Vandyke, Ondřej Dušek
Venue:
SIGDIAL
SIG:
SIGDIAL
Publisher:
Association for Computational Linguistics
Note:
Pages:
595–604
Language:
URL:
https://aclanthology.org/2022.sigdial-1.55
DOI:
10.18653/v1/2022.sigdial-1.55
Bibkey:
Cite (ACL):
Shikib Mehri, Yasemin Altun, and Maxine Eskenazi. 2022. LAD: Language Models as Data for Zero-Shot Dialog. In Proceedings of the 23rd Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 595–604, Edinburgh, UK. Association for Computational Linguistics.
Cite (Informal):
LAD: Language Models as Data for Zero-Shot Dialog (Mehri et al., SIGDIAL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.sigdial-1.55.pdf
Video:
 https://youtu.be/cdJnOFBd5mE
Data
BANKING77CLINC150HWU64STAR