The Art of Prompting: Event Detection based on Type Specific Prompts

TL;DR

This paper introduces a unified prompt-based framework for event detection that leverages event type semantics, significantly improving performance in supervised, few-shot, and zero-shot scenarios, especially with limited or no annotated data.

Contribution

It proposes a novel unified framework using event type prompts, achieving substantial performance gains over previous methods in various data-scarcity settings.

Findings

Up to 24.3% F-score improvement over baselines

Effective in few-shot and zero-shot event detection

Highlights importance of event type semantics

Abstract

We compare various forms of prompts to represent event types and develop a unified framework to incorporate the event type specific prompts for supervised, few-shot, and zero-shot event detection. The experimental results demonstrate that a well-defined and comprehensive event type prompt can significantly improve the performance of event detection, especially when the annotated data is scarce (few-shot event detection) or not available (zero-shot event detection). By leveraging the semantics of event types, our unified framework shows up to 24.3\% F-score gain over the previous state-of-the-art baselines.