Global-to-local Memory Pointer Networks for Task-Oriented Dialogue

TL;DR

This paper introduces the GLMP network, which enhances task-oriented dialogue systems by effectively integrating large, dynamic knowledge bases through a global-to-local memory mechanism, improving response accuracy and handling out-of-vocabulary issues.

Contribution

The paper presents a novel global-to-local memory pointer network that better incorporates external knowledge into dialogue generation, outperforming previous models on multiple datasets.

Findings

Improved copy accuracy and reduced out-of-vocabulary errors.

Achieved state-of-the-art results on bAbI and Stanford Multi-domain Dialogue datasets.

Enhanced handling of large, dynamic knowledge bases in dialogue systems.

Abstract

End-to-end task-oriented dialogue is challenging since knowledge bases are usually large, dynamic and hard to incorporate into a learning framework. We propose the global-to-local memory pointer (GLMP) networks to address this issue. In our model, a global memory encoder and a local memory decoder are proposed to share external knowledge. The encoder encodes dialogue history, modifies global contextual representation, and generates a global memory pointer. The decoder first generates a sketch response with unfilled slots. Next, it passes the global memory pointer to filter the external knowledge for relevant information, then instantiates the slots via the local memory pointers. We empirically show that our model can improve copy accuracy and mitigate the common out-of-vocabulary problem. As a result, GLMP is able to improve over the previous state-of-the-art models in both simulated bAbI Dialogue dataset and human-human Stanford Multi-domain Dialogue dataset on automatic and human evaluation.