English

Speak & Spell: LLM-Driven Controllable Phonetic Error Augmentation for Robust Dialogue State Tracking

Computation and Language 2025-10-31 v2 Artificial Intelligence

Abstract

Dialogue State Tracking (DST) is a key part of task-oriented dialogue systems, identifying important information in conversations. However, its accuracy drops significantly in spoken dialogue environments due to named entity errors from Automatic Speech Recognition (ASR) systems. We introduce a simple yet effective data augmentation method that targets those entities to improve the robustness of DST model. Our novel method can control the placement of errors using keyword-highlighted prompts while introducing phonetically similar errors. As a result, our method generated sufficient error patterns on keywords, leading to improved accuracy in noised and low-accuracy ASR environments.

Keywords

Cite

@article{arxiv.2409.06263,
  title  = {Speak & Spell: LLM-Driven Controllable Phonetic Error Augmentation for Robust Dialogue State Tracking},
  author = {Jihyun Lee and Solee Im and Wonjun Lee and Gary Geunbae Lee},
  journal= {arXiv preprint arXiv:2409.06263},
  year   = {2025}
}

Comments

Accepted to AACL-IJCNLP 2025

R2 v1 2026-06-28T18:39:32.373Z