Boosting Disfluency Detection with Large Language Model as Disfluency Generator

arXiv:2403.08229v1 Announce Type: new
Abstract: Current disfluency detection methods heavily rely on costly and scarce human-annotated data. To tackle this issue, some approaches employ heuristic or statistical features to generate disfluent sentences, partially improving detection performance. However, these sentences often deviate from real-life scenarios, constraining overall model enhancement. In this study, we propose a lightweight data augmentation approach for disfluency detection, utilizing the superior generative and semantic understanding capabilities of large language model (LLM) to generate disfluent sentences as augmentation …

abstract annotated data arxiv boosting cs.cl current data detection detection methods features generate generator however human issue language language model large language large language model life performance statistical type