White supremacist language classification on Siegel (test)
96.8ROC AUCWeak + Ann
Evaluation Results
| Method | Links | |
|---|---|---|
| Weak + AnnTraining source=Weakly + Manually annotated data2023.06 | 96.8 | |
| AnnotatedTraining source=Manually annotated data2023.06 | 92.9 | |
| WeakTraining source=Weakly annotated data2023.06 | 90.3 | |
| Alatawi et al. (2021)Model type=BERT2023.06 | 89.7 | |
| Weak + AnnTraining dataset=Alatawi et al. (2021) and Rieger et al. (2021)2023.06 | 88.9 | |
| WeakTraining dataset=Alatawi et al. (2021) and Rieger et al. (2021)2023.06 | 87.3 | |
| AnnotatedTraining dataset=Alatawi et al. (2021) and Rieger et al. (2021)2023.06 | 84.7 | |
| ATraining dataset=Alatawi et al. (2021)2023.06 | 83.9 | |
| Siegel et al. (2021)Method description=Dictionary match + Naive Bayes classifier2023.06 | 61.3 |