A Multichannel Convolutional Neural Network For Cross-language Dialog State Tracking

TL;DR

This paper proposes a multichannel CNN architecture for cross-language dialog state tracking, effectively handling translation errors and requiring no prior language knowledge, demonstrated on DSTC5 dataset.

Contribution

Introduction of a multichannel CNN model that improves robustness in cross-language dialog state tracking without prior language expertise.

Findings

Multichannel CNN improves robustness against translation errors.

The method performs well on DSTC5 cross-language dialog tracking.

No prior language knowledge is needed for effective tracking.

Abstract

The fifth Dialog State Tracking Challenge (DSTC5) introduces a new cross-language dialog state tracking scenario, where the participants are asked to build their trackers based on the English training corpus, while evaluating them with the unlabeled Chinese corpus. Although the computer-generated translations for both English and Chinese corpus are provided in the dataset, these translations contain errors and careless use of them can easily hurt the performance of the built trackers. To address this problem, we propose a multichannel Convolutional Neural Networks (CNN) architecture, in which we treat English and Chinese language as different input channels of one single CNN model. In the evaluation of DSTC5, we found that such multichannel architecture can effectively improve the robustness against translation errors. Additionally, our method for DSTC5 is purely machine learning based and requires no prior knowledge about the target language. We consider this a desirable property for building a tracker in the cross-language context, as not every developer will be familiar with both languages.