Enhancing Task Bot Engagement with Synthesized Open-Domain Dialog
This work addresses the need for more human-like dialog systems that integrate multiple conversation types, but it is incremental as it builds on existing dialog system research.
The paper tackled the problem of building dialog systems that fuse task-oriented and open-domain conversations by proposing a framework to automatically generate combined dialogues and a unified model called PivotBot. The result showed that PivotBot can seamlessly switch between these dialog modes, as demonstrated in evaluations.
Many efforts have been made to construct dialog systems for different types of conversations, such as task-oriented dialog (TOD) and open-domain dialog (ODD). To better mimic human-level conversations that usually fuse various dialog modes, it is essential to build a system that can effectively handle both TOD and ODD and access different knowledge sources. To address the lack of available data for the fused task, we propose a framework for automatically generating dialogues that combine knowledge-grounded ODDs and TODs in various settings. Additionally, we introduce a unified model PivotBot that is capable of appropriately adopting TOD and ODD modes and accessing different knowledge sources in order to effectively tackle the fused task. Evaluation results demonstrate the superior ability of the proposed model to switch seamlessly between TOD and ODD tasks.