A Survey on Dialog Management: Recent Advances and Challenges

TL;DR

This survey reviews recent progress and ongoing challenges in dialog management, focusing on scalability, data scarcity, and training efficiency to improve task-oriented dialog systems.

Contribution

It provides a comprehensive overview of recent advances and identifies key challenges in dialog management, guiding future research directions.

Findings

Advances in scalable dialog models

Methods to address data scarcity in dialog policy learning

Techniques to improve training efficiency for better task completion

Abstract

Dialog management (DM) is a crucial component in a task-oriented dialog system. Given the dialog history, DM predicts the dialog state and decides the next action that the dialog agent should take. Recently, dialog policy learning has been widely formulated as a Reinforcement Learning (RL) problem, and more works focus on the applicability of DM. In this paper, we survey recent advances and challenges within three critical topics for DM: (1) improving model scalability to facilitate dialog system modeling in new scenarios, (2) dealing with the data scarcity problem for dialog policy learning, and (3) enhancing the training efficiency to achieve better task-completion performance . We believe that this survey can shed a light on future research in dialog management.