Fully Statistical Neural Belief Tracking

TL;DR

This paper enhances Neural Belief Tracking for dialogue state tracking by replacing manual belief update rules with learned statistical mechanisms, enabling more adaptable and resource-efficient models across multiple languages.

Contribution

It introduces learned statistical update mechanisms into NBT, removing manual rule-based components and reducing model complexity for multilingual dialogue state tracking.

Findings

Achieves competitive performance across three languages.

Models dialogue dynamics with few additional parameters.

Provides a resource-light, robust DST framework.

Abstract

This paper proposes an improvement to the existing data-driven Neural Belief Tracking (NBT) framework for Dialogue State Tracking (DST). The existing NBT model uses a hand-crafted belief state update mechanism which involves an expensive manual retuning step whenever the model is deployed to a new dialogue domain. We show that this update mechanism can be learned jointly with the semantic decoding and context modelling parts of the NBT model, eliminating the last rule-based module from this DST framework. We propose two different statistical update mechanisms and show that dialogue dynamics can be modelled with a very small number of additional model parameters. In our DST evaluation over three languages, we show that this model achieves competitive performance and provides a robust framework for building resource-light DST models.