Zero-shot Generalization in Dialog State Tracking through Generative Question Answering
This addresses the challenge for dialog systems to adapt to constantly changing services without requiring known ontologies, representing a novel method rather than incremental improvement.
The paper tackled the problem of Dialog State Tracking (DST) generalizing to new domains and unseen slot types by introducing an ontology-free framework using generative question-answering, which improved joint goal accuracy by up to 9% over the previous state-of-the-art on the MultiWOZ 2.1 dataset in zero-shot settings.
Dialog State Tracking (DST), an integral part of modern dialog systems, aims to track user preferences and constraints (slots) in task-oriented dialogs. In real-world settings with constantly changing services, DST systems must generalize to new domains and unseen slot types. Existing methods for DST do not generalize well to new slot names and many require known ontologies of slot types and values for inference. We introduce a novel ontology-free framework that supports natural language queries for unseen constraints and slots in multi-domain task-oriented dialogs. Our approach is based on generative question-answering using a conditional language model pre-trained on substantive English sentences. Our model improves joint goal accuracy in zero-shot domain adaptation settings by up to 9% (absolute) over the previous state-of-the-art on the MultiWOZ 2.1 dataset.