Table 6 Comparison of model performance on Chinese & Portuguese datasets with previous work.

From: Adaptive ensemble techniques leveraging BERT based models for multilingual hate speech detection in Korean and english

Language/Dataset

Performance of proposed model

Model of Previous Work

Performance of Previous work

Chinese

COLD37

Acc = 0.83

F1 = 0.83

RoBERTa

Acc = 0.81

F1 = 0.81

Portuguese

ToLD-Br38

Acc = 0.76

F1 = 0.76

Brazilian Portuguese BERT, mBERT

F1 = 0.76

Back to article page