DiCoRe: Enhancing Zero-shot Event Detection via Divergent-Convergent LLM Reasoning

TL;DR

DiCoRe is a novel framework that enhances zero-shot event detection by combining divergent open-ended reasoning with convergent constrained decoding, significantly improving accuracy across multiple datasets and models.

Contribution

It introduces a divergent-convergent reasoning approach with an LLM-Judge for high-precision zero-shot event detection, outperforming prior methods.

Findings

Achieves 4-7% average F1 improvement over baselines.

Demonstrates robustness across six datasets and five domains.

Outperforms prior zero-shot, transfer-learning, and reasoning methods.

Abstract

Zero-shot Event Detection (ED), the task of identifying event mentions in natural language text without any training data, is critical for document understanding in specialized domains. Understanding the complex event ontology, extracting domain-specific triggers from the passage, and structuring them appropriately overloads and limits the utility of Large Language Models (LLMs) for zero-shot ED. To this end, we propose DiCoRe, a divergent-convergent reasoning framework that decouples the task of ED using Dreamer and Grounder. Dreamer encourages divergent reasoning through open-ended event discovery, which helps to boost event coverage. Conversely, Grounder introduces convergent reasoning to align the free-form predictions with the task-specific instructions using finite-state machine guided constrained decoding. Additionally, an LLM-Judge verifies the final outputs to ensure high precision. Through extensive experiments on six datasets across five domains and nine LLMs, we demonstrate how DiCoRe consistently outperforms prior zero-shot, transfer-learning, and reasoning baselines, achieving 4-7% average F1 gains over the best baseline -- establishing DiCoRe as a strong zero-shot ED framework.