Approaching SMM4H 2020 with Ensembles of BERT Flavours

George-Andrei Dima, Andrei-Marius Avram, Dumitru-Clementin Cercel


Abstract
This paper describes our solutions submitted to the Social Media Mining for Health Applications (#SMM4H) Shared Task 2020. We participated in the following tasks: Task 1 aimed at classifying if a tweet reports medications or not, Task 2 (only for the English dataset) aimed at discriminating if a tweet mentions adverse effects or not, and Task 5 aimed at recognizing if a tweet mentions birth defects or not. Our work focused on studying different neural network architectures based on various flavors of bidirectional Transformers (i.e., BERT), in the context of the previously mentioned classification tasks. For Task 1, we achieved an F1-score (70.5%) above the mean performance of the best scores made by all teams, whereas for Task 2, we obtained an F1-score of 37%. Also, we achieved a micro-averaged F1-score of 62% for Task 5.
Anthology ID:
2020.smm4h-1.28
Volume:
Proceedings of the Fifth Social Media Mining for Health Applications Workshop & Shared Task
Month:
December
Year:
2020
Address:
Barcelona, Spain (Online)
Editors:
Graciela Gonzalez-Hernandez, Ari Z. Klein, Ivan Flores, Davy Weissenbacher, Arjun Magge, Karen O'Connor, Abeed Sarker, Anne-Lyse Minard, Elena Tutubalina, Zulfat Miftahutdinov, Ilseyar Alimova
Venue:
SMM4H
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
153–157
Language:
URL:
https://aclanthology.org/2020.smm4h-1.28
DOI:
Bibkey:
Cite (ACL):
George-Andrei Dima, Andrei-Marius Avram, and Dumitru-Clementin Cercel. 2020. Approaching SMM4H 2020 with Ensembles of BERT Flavours. In Proceedings of the Fifth Social Media Mining for Health Applications Workshop & Shared Task, pages 153–157, Barcelona, Spain (Online). Association for Computational Linguistics.
Cite (Informal):
Approaching SMM4H 2020 with Ensembles of BERT Flavours (Dima et al., SMM4H 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-1/2020.smm4h-1.28.pdf
Data
SMM4H