OutFlip: Generating Out-of-Domain Samples for Unknown Intent Detection with Natural Language Attack

2021/05/12 by Donghyun Choi, Myeong Cheol Shin, Choi, DongHyun +5 · 1 citation
Computer Science · #Advanced Malware Detection Techniques #Computation and Language (cs.CL) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Software Engineering Research #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2105.05601

openalex publication_date 2021/05/12 · openalex created_date 2021/05/24 · openalex updated_date 2026/07/28

Abstract

Out-of-domain (OOD) input detection is vital in a task-oriented dialogue system since the acceptance of unsupported inputs could lead to an incorrect response of the system. This paper proposes OutFlip, a method to generate out-of-domain samples using only in-domain training dataset automatically. A white-box natural language attack method HotFlip is revised to generate out-of-domain samples instead of adversarial examples. Our evaluation results showed that integrating OutFlip-generated out-of-domain samples into the training dataset could significantly improve an intent classification model's out-of-domain detection performance.

Citations

• Long Short-Term Memory

Cited by

• An Interdisciplinary Review of Commonsense Reasoning and Intent Detection

Related

• Out-of-domain Detection for Natural Language Understanding in Dialog Systems
• A Hybrid Architecture for Out of Domain Intent Detection and Intent Discovery
• Out-of-Scope Intent Detection with Self-Supervision and Discriminative Training
• Diversity-grounded Channel Prototypical Learning for Out-of-Distribution Intent Detection
• Towards Textual Out-of-Domain Detection without In-Domain Labels