Boosting Naturalness of Language in Task-oriented Dialogues via Adversarial Training

Chenguang Zhu


Abstract
The natural language generation (NLG) module in a task-oriented dialogue system produces user-facing utterances conveying required information. Thus, it is critical for the generated response to be natural and fluent. We propose to integrate adversarial training to produce more human-like responses. The model uses Straight-Through Gumbel-Softmax estimator for gradient computation. We also propose a two-stage training scheme to boost performance. Empirical results show that the adversarial training can effectively improve the quality of language generation in both automatic and human evaluations. For example, in the RNN-LG Restaurant dataset, our model AdvNLG outperforms the previous state-of-the-art result by 3.6% in BLEU.
Anthology ID:
2020.sigdial-1.33
Volume:
Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue
Month:
July
Year:
2020
Address:
1st virtual meeting
Venue:
SIGDIAL
SIG:
SIGDIAL
Publisher:
Association for Computational Linguistics
Note:
Pages:
265–271
Language:
URL:
https://aclanthology.org/2020.sigdial-1.33
DOI:
Bibkey:
Cite (ACL):
Chenguang Zhu. 2020. Boosting Naturalness of Language in Task-oriented Dialogues via Adversarial Training. In Proceedings of the 21th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 265–271, 1st virtual meeting. Association for Computational Linguistics.
Cite (Informal):
Boosting Naturalness of Language in Task-oriented Dialogues via Adversarial Training (Zhu, SIGDIAL 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/paclic-22-ingestion/2020.sigdial-1.33.pdf
Video:
 https://youtube.com/watch?v=JZHzDvmG6Ns