Query and Extract: Refining Event Extraction as Type-oriented Binary Decoding
This work addresses the limitation of pre-defined event types in extraction tasks, benefiting researchers and practitioners in natural language processing by enabling more flexible and unified modeling across ontologies.
The paper tackles the problem of event extraction by reframing it as a type-oriented binary decoding task, using natural language queries for event types and argument roles, and achieves state-of-the-art performance on ACE and ERE datasets with significant improvements in zero-shot extraction.
Event extraction is typically modeled as a multi-class classification problem where event types and argument roles are treated as atomic symbols. These approaches are usually limited to a set of pre-defined types. We propose a novel event extraction framework that uses event types and argument roles as natural language queries to extract candidate triggers and arguments from the input text. With the rich semantics in the queries, our framework benefits from the attention mechanisms to better capture the semantic correlation between the event types or argument roles and the input text. Furthermore, the query-and-extract formulation allows our approach to leverage all available event annotations from various ontologies as a unified model. Experiments on ACE and ERE demonstrate that our approach achieves state-of-the-art performance on each dataset and significantly outperforms existing methods on zero-shot event extraction.