Estimating Soft Labels for Out-of-Domain Intent Detection

TL;DR

This paper introduces ASoul, an adaptive soft pseudo labeling method that improves out-of-domain intent detection by estimating soft labels for pseudo OOD samples using an embedding graph and co-training, leading to better detection performance.

Contribution

The paper proposes a novel adaptive soft pseudo labeling approach with an embedding graph and co-training framework for enhanced OOD intent detection.

Findings

ASoul outperforms baseline methods on three benchmark datasets.

Soft labeling reduces noise from pseudo OOD samples.

The method improves OOD detection accuracy significantly.

Abstract

Out-of-Domain (OOD) intent detection is important for practical dialog systems. To alleviate the issue of lacking OOD training samples, some works propose synthesizing pseudo OOD samples and directly assigning one-hot OOD labels to these pseudo samples. However, these one-hot labels introduce noises to the training process because some hard pseudo OOD samples may coincide with In-Domain (IND) intents. In this paper, we propose an adaptive soft pseudo labeling (ASoul) method that can estimate soft labels for pseudo OOD samples when training OOD detectors. Semantic connections between pseudo OOD samples and IND intents are captured using an embedding graph. A co-training framework is further introduced to produce resulting soft labels following the smoothness assumption, i.e., close samples are likely to have similar labels. Extensive experiments on three benchmark datasets show that ASoul consistently improves the OOD detection performance and outperforms various competitive baselines.