Dialogue Summaries as Dialogue States (DS2), Template-Guided Summarization for Few-shot Dialogue State Tracking

TL;DR

This paper introduces DS2, a novel approach that reformulates dialogue state tracking as a dialogue summarization task using template-guided synthetic summaries, achieving superior few-shot performance and faster inference.

Contribution

The paper proposes a new method that treats dialogue state tracking as summarization, leveraging synthetic templates to improve few-shot learning and speed.

Findings

Outperforms previous few-shot DST methods on MultiWoZ datasets

Enables rapid training and inference by generating all states simultaneously

Template naturalness significantly impacts training success

Abstract

Annotating task-oriented dialogues is notorious for the expensive and difficult data collection process. Few-shot dialogue state tracking (DST) is a realistic solution to this problem. In this paper, we hypothesize that dialogue summaries are essentially unstructured dialogue states; hence, we propose to reformulate dialogue state tracking as a dialogue summarization problem. To elaborate, we train a text-to-text language model with synthetic template-based dialogue summaries, generated by a set of rules from the dialogue states. Then, the dialogue states can be recovered by inversely applying the summary generation rules. We empirically show that our method DS2 outperforms previous works on few-shot DST in MultiWoZ 2.0 and 2.1, in both cross-domain and multi-domain settings. Our method also exhibits vast speedup during both training and inference as it can generate all states at once. Finally, based on our analysis, we discover that the naturalness of the summary templates plays a key role for successful training.