TOD-Flow: Modeling the Structure of Task-Oriented Dialogues

Sungryull Sohn, Yiwei Lyu, Anthony Liu, Lajanugen Logeswaran, Dong-Ki Kim, Dongsub Shim, Honglak Lee


Abstract
Task-Oriented Dialogue (TOD) systems have become crucial components in interactive artificial intelligence applications. While recent advances have capitalized on pre-trained language models (PLMs), they exhibit limitations regarding transparency and controllability. To address these challenges, we propose a novel approach focusing on inferring the TOD-flow graph from dialogue data annotated with dialog acts, uncovering the underlying task structure in the form of a graph. The inferred TOD-flow graph can be easily integrated with any dialogue model to improve its prediction performance, transparency, and controllability. Our TOD-flow graph learns what a model can, should, and should not predict, effectively reducing the search space and providing a rationale for the model’s prediction. We show that the proposed TOD-flow graph better resemble human-annotated graphs compared to prior approaches. Furthermore, when combined with several dialogue policies and end-to-end dialogue models, we demonstrate that our approach significantly improves dialog act classification and end-to-end response generation performance in the MultiWOZ and SGD benchmarks.
Anthology ID:
2023.emnlp-main.204
Volume:
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3355–3371
Language:
URL:
https://aclanthology.org/2023.emnlp-main.204
DOI:
10.18653/v1/2023.emnlp-main.204
Bibkey:
Cite (ACL):
Sungryull Sohn, Yiwei Lyu, Anthony Liu, Lajanugen Logeswaran, Dong-Ki Kim, Dongsub Shim, and Honglak Lee. 2023. TOD-Flow: Modeling the Structure of Task-Oriented Dialogues. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 3355–3371, Singapore. Association for Computational Linguistics.
Cite (Informal):
TOD-Flow: Modeling the Structure of Task-Oriented Dialogues (Sohn et al., EMNLP 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.emnlp-main.204.pdf
Video:
 https://aclanthology.org/2023.emnlp-main.204.mp4