Ravenclaw: dialog management using hierarchical task decomposition and an expectation agenda

TLDR

RavenClaw is a new dialog management framework succeeding the Agenda architecture, aimed at simplifying the development of spoken dialog systems. RavenClaw focuses on task specification, with its engine automatically generating domain‑independent conversational behaviors. RavenClaw separates task from discourse behavior, enabling rapid component development and has been successfully applied to five domains, indicating its generality.

Abstract

We describe RavenClaw, a new dialog management framework developed as a successor to the Agenda [1] architecture used in the CMU Communicator. RavenClaw introduces a clear separation between task and discourse behavior specification, and allows rapid development of dialog management components for spoken dialog systems operating in complex, goal-oriented domains. The system development effort is focused entirely on the specification of the dialog task, while a rich set of domain-independent conversational behaviors are transparently generated by the dialog engine. To date, RavenClaw has been applied to five different domains allowing us to draw some preliminary conclusions as to the generality of the approach. We briefly describe our experience in developing these systems.