PUM at SemEval-2020 Task 12: Aggregation of Transformer-based Models’ Features for Offensive Language Recognition

Piotr Janiszewski, Mateusz Skiba, Urszula Walińska


Abstract
In this paper, we describe the PUM team’s entry to the SemEval-2020 Task 12. Creating our solution involved leveraging two well-known pretrained models used in natural language processing: BERT and XLNet, which achieve state-of-the-art results in multiple NLP tasks. The models were fine-tuned for each subtask separately and features taken from their hidden layers were combinedand fed into a fully connected neural network. The model using aggregated Transformer featurescan serve as a powerful tool for offensive language identification problem. Our team was ranked7th out of 40 in Sub-task C - Offense target identification with 64.727% macro F1-score and 64thout of 85 in Sub-task A - Offensive language identification (89.726% F1-score).
Anthology ID:
2020.semeval-1.210
Volume:
Proceedings of the Fourteenth Workshop on Semantic Evaluation
Month:
December
Year:
2020
Address:
Barcelona (online)
Editors:
Aurelie Herbelot, Xiaodan Zhu, Alexis Palmer, Nathan Schneider, Jonathan May, Ekaterina Shutova
Venue:
SemEval
SIG:
SIGLEX
Publisher:
International Committee for Computational Linguistics
Note:
Pages:
1615–1621
Language:
URL:
https://aclanthology.org/2020.semeval-1.210
DOI:
10.18653/v1/2020.semeval-1.210
Bibkey:
Cite (ACL):
Piotr Janiszewski, Mateusz Skiba, and Urszula Walińska. 2020. PUM at SemEval-2020 Task 12: Aggregation of Transformer-based Models’ Features for Offensive Language Recognition. In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 1615–1621, Barcelona (online). International Committee for Computational Linguistics.
Cite (Informal):
PUM at SemEval-2020 Task 12: Aggregation of Transformer-based Models’ Features for Offensive Language Recognition (Janiszewski et al., SemEval 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.semeval-1.210.pdf
Data
OLID