Knowledge-grounded Dialog State Tracking
This work addresses the challenge of adapting dialog systems to new tasks and domains with different schemas, which is an incremental improvement for natural language processing applications.
The paper tackles the problem of inefficient training and poor transferability in dialog state tracking by proposing a method that grounds predictions on externally encoded knowledge rather than implicitly encoding domain-specific knowledge into model parameters. The method demonstrates superior performance over strong baselines, particularly in few-shot learning settings.
Knowledge (including structured knowledge such as schema and ontology, and unstructured knowledge such as web corpus) is a critical part of dialog understanding, especially for unseen tasks and domains. Traditionally, such domain-specific knowledge is encoded implicitly into model parameters for the execution of downstream tasks, which makes training inefficient. In addition, such models are not easily transferable to new tasks with different schemas. In this work, we propose to perform dialog state tracking grounded on knowledge encoded externally. We query relevant knowledge of various forms based on the dialog context where such information can ground the prediction of dialog states. We demonstrate superior performance of our proposed method over strong baselines, especially in the few-shot learning setting.