KEIS@JUST at SemEval-2020 Task 12: Identifying Multilingual Offensive Tweets Using Weighted Ensemble and Fine-Tuned BERT

Saja Tawalbeh, Mahmoud Hammad, Mohammad AL-Smadi


Abstract
This research presents our team KEIS@JUST participation at SemEval-2020 Task 12 which represents shared task on multilingual offensive language. We participated in all the provided languages for all subtasks except sub-task-A for the English language. Two main approaches have been developed the first is performed to tackle both languages Arabic and English, a weighted ensemble consists of Bi-GRU and CNN followed by Gaussian noise and global pooling layer multiplied by weights to improve the overall performance. The second is performed for other languages, a transfer learning from BERT beside the recurrent neural networks such as Bi-LSTM and Bi-GRU followed by a global average pooling layer. Word embedding and contextual embedding have been used as features, moreover, data augmentation has been used only for the Arabic language.
Anthology ID:
2020.semeval-1.269
Volume:
Proceedings of the Fourteenth Workshop on Semantic Evaluation
Month:
December
Year:
2020
Address:
Barcelona (online)
Venues:
COLING | SemEval
SIGs:
SIGLEX | SIGSEM
Publisher:
International Committee for Computational Linguistics
Note:
Pages:
2035–2044
Language:
URL:
https://aclanthology.org/2020.semeval-1.269
DOI:
10.18653/v1/2020.semeval-1.269
Bibkey:
Cite (ACL):
Saja Tawalbeh, Mahmoud Hammad, and Mohammad AL-Smadi. 2020. KEIS@JUST at SemEval-2020 Task 12: Identifying Multilingual Offensive Tweets Using Weighted Ensemble and Fine-Tuned BERT. In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 2035–2044, Barcelona (online). International Committee for Computational Linguistics.
Cite (Informal):
KEIS@JUST at SemEval-2020 Task 12: Identifying Multilingual Offensive Tweets Using Weighted Ensemble and Fine-Tuned BERT (Tawalbeh et al., SemEval 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/update-css-js/2020.semeval-1.269.pdf
Data
OLID