Efficient Dialogue State Tracking by Masked Hierarchical Transformer
This addresses the problem of building cross-lingual dialogue systems for low-resource languages, though it appears incremental as it builds on existing DSTC challenge frameworks.
The paper tackles cross-lingual multi-domain dialogue state tracking by proposing a method that jointly learns slot operation classification and state tracking with a novel mask mechanism for contextual fusion, achieving joint accuracies of 62.37% on MultiWOZ and 23.96% on CrossWOZ datasets.
This paper describes our approach to DSTC 9 Track 2: Cross-lingual Multi-domain Dialog State Tracking, the task goal is to build a Cross-lingual dialog state tracker with a training set in rich resource language and a testing set in low resource language. We formulate a method for joint learning of slot operation classification task and state tracking task respectively. Furthermore, we design a novel mask mechanism for fusing contextual information about dialogue, the results show the proposed model achieves excellent performance on DSTC Challenge II with a joint accuracy of 62.37% and 23.96% in MultiWOZ(en - zh) dataset and CrossWOZ(zh - en) dataset, respectively.