Adaptive Schema-aware Event Extraction with Retrieval-Augmented Generation

Event extraction (EE) is a fundamental task in natural language processing (NLP) that involves identifying and extracting event information from unstructured text. Effective EE in real-world scenarios requires two key steps: selecting appropriate schemas from hundreds of candidates and executing the extraction process. Existing research exhibits two critical gaps: (1) the rigid schema fixation in existing pipeline systems, and (2) the absence of benchmarks for evaluating joint schema matching and extraction. While large language models (LLMs) show promise, challenges such as schema hallucination and the inefficient use of long contexts for schema selection hinder their practical deployment. In response, we propose Adaptive Schemaaware Event Extraction (ASEE), a retrievalaugmented generation paradigm. ASEE first enriches schemas via schema paraphrasing to improve retrieval accuracy and then performs targeted, schema-guided extraction. To facilitate rigorous evaluation, we construct the Multi-Dimensional Schema-aware Event Extraction (MD-SEE) benchmark, which systematically consolidates 12 datasets across diverse domains, complexity levels, and language settings. Extensive evaluations on MD-SEE show that our proposed ASEE demonstrates strong adaptability across various scenarios, significantly improving the performance of joint schema retrieval and event extraction. Our codes and datasets are available at https:// github.com/USTC-StarTeam/ASEE.git