BERT of all trades, master of some

Denis Gordeev, Olga Lykova


Abstract
This paper describes our results for TRAC 2020 competition held together with the conference LREC 2020. Our team name was Ms8qQxMbnjJMgYcw. The competition consisted of 2 subtasks in 3 languages (Bengali, English and Hindi) where the participants’ task was to classify aggression in short texts from social media and decide whether it is gendered or not. We used a single BERT-based system with two outputs for all tasks simultaneously. Our model placed first in English and second in Bengali gendered text classification competition tasks with 0.87 and 0.93 in F1-score respectively.
Anthology ID:
2020.trac-1.15
Volume:
Proceedings of the Second Workshop on Trolling, Aggression and Cyberbullying
Month:
May
Year:
2020
Address:
Marseille, France
Editors:
Ritesh Kumar, Atul Kr. Ojha, Bornini Lahiri, Marcos Zampieri, Shervin Malmasi, Vanessa Murdock, Daniel Kadar
Venue:
TRAC
SIG:
Publisher:
European Language Resources Association (ELRA)
Note:
Pages:
93–98
Language:
English
URL:
https://aclanthology.org/2020.trac-1.15
DOI:
Bibkey:
Cite (ACL):
Denis Gordeev and Olga Lykova. 2020. BERT of all trades, master of some. In Proceedings of the Second Workshop on Trolling, Aggression and Cyberbullying, pages 93–98, Marseille, France. European Language Resources Association (ELRA).
Cite (Informal):
BERT of all trades, master of some (Gordeev & Lykova, TRAC 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.trac-1.15.pdf