English

Simulating Task-Oriented Dialogues with State Transition Graphs and Large Language Models

Computation and Language 2024-04-24 v1

Abstract

This paper explores SynTOD, a new synthetic data generation approach for developing end-to-end Task-Oriented Dialogue (TOD) Systems capable of handling complex tasks such as intent classification, slot filling, conversational question-answering, and retrieval-augmented response generation, without relying on crowdsourcing or real-world data. SynTOD utilizes a state transition graph to define the desired behavior of a TOD system and generates diverse, structured conversations through random walks and response simulation using large language models (LLMs). In our experiments, using graph-guided response simulations leads to significant improvements in intent classification, slot filling and response relevance compared to naive single-prompt simulated conversations. We also investigate the end-to-end TOD effectiveness of different base and instruction-tuned LLMs, with and without the constructed synthetic conversations. Finally, we explore how various LLMs can evaluate responses in a TOD system and how well they are correlated with human judgments. Our findings pave the path towards quick development and evaluation of domain-specific TOD systems. We release our datasets, models, and code for research purposes.

Keywords

Cite

@article{arxiv.2404.14772,
  title  = {Simulating Task-Oriented Dialogues with State Transition Graphs and Large Language Models},
  author = {Chris Samarinas and Pracha Promthaw and Atharva Nijasure and Hansi Zeng and Julian Killingback and Hamed Zamani},
  journal= {arXiv preprint arXiv:2404.14772},
  year   = {2024}
}
R2 v1 2026-06-28T16:03:13.751Z