Novel Slot Detection: A Benchmark for Discovering Unknown Slot Types in the Task-Oriented Dialogue System
This work tackles the problem of out-of-domain slot detection for task-oriented dialogue systems, providing a foundational benchmark but is incremental as it builds on existing slot filling frameworks.
The paper introduces the Novel Slot Detection (NSD) task to identify unknown slot types in task-oriented dialogue systems, addressing the limitation of existing models that only recognize pre-defined slots, and establishes a benchmark with datasets and baselines for future research.
Existing slot filling models can only recognize pre-defined in-domain slot types from a limited slot set. In the practical application, a reliable dialogue system should know what it does not know. In this paper, we introduce a new task, Novel Slot Detection (NSD), in the task-oriented dialogue system. NSD aims to discover unknown or out-of-domain slot types to strengthen the capability of a dialogue system based on in-domain training data. Besides, we construct two public NSD datasets, propose several strong NSD baselines, and establish a benchmark for future work. Finally, we conduct exhaustive experiments and qualitative analysis to comprehend key challenges and provide new guidance for future directions.