APP: Adaptive Prototypical Pseudo-Labeling for Few-shot OOD Detection
This addresses a practical challenge for task-oriented dialogue systems by enabling OOD detection with scarce labeled data, though it appears incremental as it builds on existing OOD detection methods.
The paper tackles the problem of detecting out-of-domain (OOD) intents in task-oriented dialogue systems under a few-shot setting with limited labeled in-domain (IND) data and massive unlabeled mixed data, proposing an adaptive prototypical pseudo-labeling (APP) method that demonstrates effectiveness in experiments.
Detecting out-of-domain (OOD) intents from user queries is essential for a task-oriented dialogue system. Previous OOD detection studies generally work on the assumption that plenty of labeled IND intents exist. In this paper, we focus on a more practical few-shot OOD setting where there are only a few labeled IND data and massive unlabeled mixed data that may belong to IND or OOD. The new scenario carries two key challenges: learning discriminative representations using limited IND data and leveraging unlabeled mixed data. Therefore, we propose an adaptive prototypical pseudo-labeling (APP) method for few-shot OOD detection, including a prototypical OOD detection framework (ProtoOOD) to facilitate low-resource OOD detection using limited IND data, and an adaptive pseudo-labeling method to produce high-quality pseudo OOD\&IND labels. Extensive experiments and analysis demonstrate the effectiveness of our method for few-shot OOD detection.