Structured belief states are crucial for user goal tracking and database query in task-oriented
dialog systems. However, training belief trackers often requires expensive turn-level …

Existing dialogue state tracking (DST) models require plenty of labeled data. However,
collecting high-quality labels is costly, especially when the number of domains increases. In …

In this paper, we propose a meta-learning based semi-supervised explicit dialogue state
tracker (SEDST) for neural dialogue generation, denoted as MEDST. Our main motivation is …

Most existing intent discovery methods leverage representation learning and clustering to
transfer the prior knowledge of known intents to unknown ones. The learned representations …

Multimodal machine learning (MMML) is a vibrant multi-disciplinary research field which
addresses some of the original goals of artificial intelligence by integrating and modelling …

This paper presents a predictive study on the progress of conversations. Specifically, we
estimate the residual life for conversations, which is defined as the count of new turns to …

In view of the widespread use of social platforms, interpersonal communications have come
to play an increasingly crucial role in our daily activities. Nevertheless, although every …

The success of deep learning methods has stimulated the rapid development of many NLP
research areas. Still, task-oriented dialogue modelling remains challenging due to both the …