Targeted Augmentation for Low-Resource Event Extraction

arXiv:2405.08729v1 Announce Type: new
Abstract: Addressing the challenge of low-resource information extraction remains an ongoing issue due to the inherent information scarcity within limited training examples. Existing data augmentation methods, considered potential solutions, struggle to strike a balance between weak augmentation (e.g., synonym augmentation) and drastic augmentation (e.g., conditional generation without proper guidance). This paper introduces a novel paradigm that employs targeted augmentation and back validation to produce augmented examples with enhanced diversity, polarity, accuracy, and coherence. Extensive experimental results …

abstract arxiv augmentation balance challenge cs.ai cs.cl data event examples extraction guidance information information extraction issue low solutions strike struggle training type