Views
No views yet
| Metric | Score |
|---|---|
| Accuracy | 81.3% |
| F1-Macro | 0.810 |
| F1-Weighted | 0.813 |
| MCC | 0.621 |
| Class | Precision | Recall | F1-Score |
|---|---|---|---|
| ACC (Acceptable) | 0.777 | 0.803 | 0.790 |
| OFF (Offensive) | 0.842 | 0.820 | 0.831 |
1from src.models.bertic import BERTicTrainer
2
3# Load model
4trainer = BERTicTrainer()
5trainer.load("path/to/model")
6
7# Predict
8texts = ["Ovo je normalan komentar.", "Svi su oni lopovi!"]
9predictions = trainer.predict(texts)
10print(predictions) # ['ACC', 'OFF']ACC - Acceptable: No offensive contentOFF - Offensive: Contains hate speech, insults, or inappropriate content1@misc{croatian-hate-speech-2026,
2 author = {Jurić, Duje and Matošević, Teo and Radolović, Teo},
3 title = {Detection of Hate Speech on Croatian Online Portals Using NLP Methods},
4 year = {2026},
5 publisher = {University of Zagreb, FER},
6 url = {https://github.com/TeoMatosevic/slur-analysis-model}
7}