fbpx

EEUM researcher wins Best Student Paper Award

Luís Filipe Cunha, Ricardo Campos and Alípio Jorge won the Best Student Paper Award at the 22nd Portuguese Conference on Artificial Intelligence (EPIA’23), with the article “Event Extraction for Portuguese: A QA-driven Approach using ACE-2005”.

EPIA’23 took place between 5th and 8th September 2023 in the Azores, on the island of Faial. This conference is organised by the Portuguese Association for Artificial Intelligence (APPIA).

The work presented in the article proposes a new method for extracting events from Portuguese texts, which involves two stages: (1) classifying and identifying the centre word of an event (trigger); (2) extracting event arguments using an extractive Q&A model. In order to produce models capable of performing these tasks, the authors trained models based on BERTimbau using the SQuAD and ACE-2005 datasets, the latter being a reference in the area of event extraction. Using these models, they developed an event extraction framework for the Portuguese language.

This article will subsequently be published in Springer’s Lecture Notes in Artificial Intelligence (LNAI).

Authors and affiliations:

Luís Filipe Cunha – University of Minho (visiting lecturer at the Department of Computer Science of the School of Engineering) and LIAAD-INESC TEC, FCUP – University of Porto (PhD student);

Ricardo Campos – LIAAD-INESC TEC, University of Beira Interior;

Alípio Jorge – LIAAD-INESC TEC, FCUP – University of Porto.

 

BSPAward