SPARTA at CASE 2021 Task 1: Evaluating Different Techniques to Improve Event Extraction
Arthur Müller, Andreas Dafnos · 2022
We participated in the Shared Task 1 at CASE 2021, Subtask 4 on protest event extraction from news articles (Hürriyetoglu et al., 2022) and examined different techniques aimed at improving the performance of the winning system from the last competition round (Hürriyetoglu et al., 2021).We evaluated in-domain pretraining, task-specific pre-fine-tuning, alternative loss function, translation of the English training dataset into other target languages (i.e., Portuguese, Spanish, and Hindi) for the token classification task, and a simple data augmentation technique by random sentence reordering.This paper summarizes the results, showing that random sentence reordering leads to a consistent improvement of the model performance.