Retrieve & Memorize: Dialog Policy Learning with Multi-Action Memory

TL;DR

This paper introduces a retrieve-and-memorize framework for dialogue policy learning that improves system action generation by retrieving relevant actions and adaptively selecting key information, leading to competitive results on multi-domain datasets.

Contribution

The paper proposes a novel retrieve-and-memorize approach with a neural retrieval module and memory-augmented multi-decoder for better dialogue policy learning.

Findings

Achieves competitive performance on MultiWOZ datasets.

Effectively retrieves and utilizes multiple candidate actions.

Improves generation of system responses in task-oriented dialogues.

Abstract

Dialogue policy learning, a subtask that determines the content of system response generation and then the degree of task completion, is essential for task-oriented dialogue systems. However, the unbalanced distribution of system actions in dialogue datasets often causes difficulty in learning to generate desired actions and responses. In this paper, we propose a retrieve-and-memorize framework to enhance the learning of system actions. Specially, we first design a neural context-aware retrieval module to retrieve multiple candidate system actions from the training set given a dialogue context. Then, we propose a memory-augmented multi-decoder network to generate the system actions conditioned on the candidate actions, which allows the network to adaptively select key information in the candidate actions and ignore noises. We conduct experiments on the large-scale multi-domain task-oriented dialogue dataset MultiWOZ 2.0 and MultiWOZ 2.1. Experimental results show that our method achieves competitive performance among several state-of-the-art models in the context-to-response generation task.