Plan, Generate and Complicate: Improving Low-resource Dialogue State Tracking via Easy-to-Difficult Zero-shot Data Augmentation

TL;DR

Low-resource dialogue state tracking suffers from limited annotated data and the need to handle complex co-reference relations. The authors introduce EDZ-DA, an easy-to-difficult zero-shot data augmentation framework that leverages large language models for domain-relations planning, dialogue-flow generation, and co-reference-aware data complication, complemented by slot-value permutation. The approach yields substantial improvements over strong baselines on MultiWOZ, especially in co-reference-slot tracking, and even rivals much larger models in some settings. This work offers a practical, scalable path to improve DST for low-resource ToD systems and highlights the importance of data complexity and structured labeling in augmentation pipelines.

Abstract

Data augmentation methods have been a promising direction to improve the performance of small models for low-resource dialogue state tracking. However, traditional methods rely on pre-defined user goals and neglect the importance of data complexity in this task. In this paper, we propose EDZ-DA, an Easy-to-Difficult Zero-shot Data Augmentation framework for low-resource dialogue state tracking that utilizes large language models to automatically catch the relationships of different domains and then generate the dialogue data. We also complicate the dialogues based on the domain relation to enhance the model's capability for co-reference slot tracking. Furthermore, we permute slot values to mitigate the influence of output orders and the problem of incomplete value generation. Experimental results illustrate the superiority of our proposed method compared to previous strong data augmentation baselines on MultiWOZ.

Figures (2)

• Figure 1: Differences between traditional methods and our method.
• Figure 2: The overview of our proposed data augmentation method.