Unsupervised Learning of Hierarchical Conversation Structure
This work addresses the challenge of domain-dependent conversation structure for applications in automatic understanding and summarization, though it appears incremental as it builds on existing neural models.
The paper tackles the problem of automatically understanding and summarizing human conversations by introducing an unsupervised approach to learn hierarchical conversation structure, including turn and sub-dialogue labels, which enhances neural models for three conversation-level understanding tasks.
Human conversations can evolve in many different ways, creating challenges for automatic understanding and summarization. Goal-oriented conversations often have meaningful sub-dialogue structure, but it can be highly domain-dependent. This work introduces an unsupervised approach to learning hierarchical conversation structure, including turn and sub-dialogue segment labels, corresponding roughly to dialogue acts and sub-tasks, respectively. The decoded structure is shown to be useful in enhancing neural models of language for three conversation-level understanding tasks. Further, the learned finite-state sub-dialogue network is made interpretable through automatic summarization.