Prompt Learning for Domain Adaptation in Task-Oriented Dialogue
This addresses the barrier for conversation designers in creating natural, user-friendly dialogue systems by reducing development effort, though it appears incremental as it builds on existing prompt learning and domain adaptation techniques.
The paper tackles the problem of high complexity and cost in developing task-oriented dialogue systems by proposing canonical forms as a lightweight semantic representation for intent classification, showing that tuning soft prompts for a frozen large language model enables zero- or few-shot generalization to new domains with sample efficiency.
Conversation designers continue to face significant obstacles when creating production quality task-oriented dialogue systems. The complexity and cost involved in schema development and data collection is often a major barrier for such designers, limiting their ability to create natural, user-friendly experiences. We frame the classification of user intent as the generation of a canonical form, a lightweight semantic representation using natural language. We show that canonical forms offer a promising alternative to traditional methods for intent classification. By tuning soft prompts for a frozen large language model, we show that canonical forms generalize very well to new, unseen domains in a zero- or few-shot setting. The method is also sample-efficient, reducing the complexity and effort of developing new task-oriented dialogue domains.