Events Beyond ACE: Curated Training for Events
This work addresses annotation efficiency and performance in event extraction for NLP researchers, offering an incremental improvement over existing methods like ACE.
The paper tackled the problem of event argument extraction by introducing curated training, a human-driven annotation approach where annotators use interactive search to find informative text snippets for teaching the system, resulting in a 6% reduction in error compared to traditional methods and faster learning curves for expert annotators.
We explore a human-driven approach to annotation, curated training (CT), in which annotation is framed as teaching the system by using interactive search to identify informative snippets of text to annotate, unlike traditional approaches which either annotate preselected text or use active learning. A trained annotator performed 80 hours of CT for the thirty event types of the NIST TAC KBP Event Argument Extraction evaluation. Combining this annotation with ACE results in a 6% reduction in error and the learning curve of CT plateaus more slowly than for full-document annotation. 3 NLP researchers performed CT for one event type and showed much sharper learning curves with all three exceeding ACE performance in less than ninety minutes, suggesting that CT can provide further benefits when the annotator deeply understands the system.