@inproceedings{kaiser-etal-2024-learning,
    title = "Learning from Relevant Subgoals in Successful Dialogs using Iterative Training for Task-oriented Dialog Systems",
    author = {Kaiser, Magdalena  and
      Ernst, Patrick  and
      Szarvas, Gy{\"o}rgy},
    editor = "Al-Onaizan, Yaser  and
      Bansal, Mohit  and
      Chen, Yun-Nung",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
    month = nov,
    year = "2024",
    address = "Miami, Florida, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2024.findings-emnlp.362/",
    doi = "10.18653/v1/2024.findings-emnlp.362",
    pages = "6236--6246",
    abstract = "Task-oriented Dialog (ToD) systems have to solve multiple subgoals to accomplish user goals, whereas feedback is often obtained only at the end of the dialog. In this work, we propose SUIT (SUbgoal-aware ITerative Training), an iterative training approach for improving ToD systems. We sample dialogs from the model we aim to improve and determine subgoals that contribute to dialog success using distant supervision to obtain high quality training samples. We show how this data improves supervised fine-tuning or, alternatively, preference learning results. Performance improves when applying these steps over several iterations: SUIT reaches new state-of-the-art performance on a popular ToD benchmark."
}