Fine-Grained Temporal Relation Extraction
This work addresses the need for fine-grained temporal analysis in natural language processing, offering a novel dataset and framework for researchers in this domain.
The paper tackled the problem of modeling temporal relations and event durations by introducing a semantic framework that maps event pairs to real-valued scales, and constructed the largest temporal relations dataset from the Universal Dependencies English Web Treebank, achieving strong results in joint prediction tasks.
We present a novel semantic framework for modeling temporal relations and event durations that maps pairs of events to real-valued scales. We use this framework to construct the largest temporal relations dataset to date, covering the entirety of the Universal Dependencies English Web Treebank. We use this dataset to train models for jointly predicting fine-grained temporal relations and event durations. We report strong results on our data and show the efficacy of a transfer-learning approach for predicting categorical relations.