Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems
arXiv:2109.08820
Abstract
Most prior work on task-oriented dialogue systems is restricted to supporting domain APIs. However, users may have requests that are out of the scope of these APIs. This work focuses on identifying such user requests. Existing methods for this task mainly rely on fine-tuning pre-trained models on large annotated data. We propose a novel method, REDE, based on adaptive representation learning and density estimation. REDE can be applied to zero-shot cases, and quickly learns a high-performing detector with only a few shots by updating less than 3K parameters. We demonstrate REDE's competitive performance on DSTC9 data and our newly collected test set.
To appear at NLP4ConvAI workshop of EMNLP 2021
References in corpus (4)
- Deep Anomaly Detection with Outlier Exposure
- Whitening Sentence Representations for Better Semantics and Faster Retrieval
- From Machine Reading Comprehension to Dialogue State Tracking: Bridging the Gap
- Beyond Domain APIs: Task-oriented Conversational Modeling with Unstructured Knowledge Access Track in DSTC9