Schema-Guided Paradigm for Zero-Shot Dialog

TL;DR

This paper introduces the Schema Attention Model (SAM) that explicitly uses schema-guided information to enable zero-shot dialog task transfer, significantly improving performance on unseen tasks.

Contribution

The paper proposes a novel schema-guided paradigm and SAM model that enhance zero-shot dialog system adaptation to unseen tasks and domains.

Findings

SAM achieves +22 F1 score over prior work in zero-shot settings.

Schema representations improve dialog policy generalization.

Ablation studies confirm the effectiveness of SAM.

Abstract

Developing mechanisms that flexibly adapt dialog systems to unseen tasks and domains is a major challenge in dialog research. Neural models implicitly memorize task-specific dialog policies from the training data. We posit that this implicit memorization has precluded zero-shot transfer learning. To this end, we leverage the schema-guided paradigm, wherein the task-specific dialog policy is explicitly provided to the model. We introduce the Schema Attention Model (SAM) and improved schema representations for the STAR corpus. SAM obtains significant improvement in zero-shot settings, with a +22 F1 score improvement over prior work. These results validate the feasibility of zero-shot generalizability in dialog. Ablation experiments are also presented to demonstrate the efficacy of SAM.