Beyond Pairwise: Global Zero-shot Temporal Graph Generation
This addresses the computational inefficiency and lack of global consistency in TRE for NLP applications, though it is incremental as it builds on existing zero-shot and supervised methods.
The paper tackles the problem of temporal relation extraction (TRE) by proposing a zero-shot method that generates a complete temporal graph in one step and refines it with constraint optimization, outperforming existing zero-shot approaches and offering a competitive alternative to supervised models.
Temporal relation extraction (TRE) is a fundamental task in natural language processing (NLP) that involves identifying the temporal relationships between events in a document. Despite the advances in large language models (LLMs), their application to TRE remains limited. Most existing approaches rely on pairwise classification, where event pairs are classified in isolation, leading to computational inefficiency and a lack of global consistency in the resulting temporal graph. In this work, we propose a novel zero-shot method for TRE that generates a document's complete temporal graph in a single step, followed by temporal constraint optimization to refine predictions and enforce temporal consistency across relations. Additionally, we introduce OmniTemp, a new dataset with complete annotations for all pairs of targeted events within a document. Through experiments and analyses, we demonstrate that our method outperforms existing zero-shot approaches and offers a competitive alternative to supervised TRE models.