Extracting Entities and Events as a Single Task Using a Transition-Based Neural Model
Abstract
The task of event extraction contains subtasks including detections for entity mentions, event triggers and argument roles. Traditional methods solve them as a pipeline, which does not make use of task correlation for their mutual benefits. There have been recent efforts towards building a joint model for all tasks. However, due to technical challenges, there has not been work predicting the joint output structure as a single task. We build a first model to this end using a neural transition-based framework, incrementally predicting complex joint structures in a state-transition process. Results on standard benchmarks show the benefits of the joint model, which gives the best result in the literature.
Cite
Text
Zhang et al. "Extracting Entities and Events as a Single Task Using a Transition-Based Neural Model." International Joint Conference on Artificial Intelligence, 2019. doi:10.24963/IJCAI.2019/753Markdown
[Zhang et al. "Extracting Entities and Events as a Single Task Using a Transition-Based Neural Model." International Joint Conference on Artificial Intelligence, 2019.](https://mlanthology.org/ijcai/2019/zhang2019ijcai-extracting/) doi:10.24963/IJCAI.2019/753BibTeX
@inproceedings{zhang2019ijcai-extracting,
title = {{Extracting Entities and Events as a Single Task Using a Transition-Based Neural Model}},
author = {Zhang, Junchi and Qin, Yanxia and Zhang, Yue and Liu, Mengchi and Ji, Donghong},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2019},
pages = {5422-5428},
doi = {10.24963/IJCAI.2019/753},
url = {https://mlanthology.org/ijcai/2019/zhang2019ijcai-extracting/}
}