Hate Speech Classifier for Social Media Content in Slovenian Language

A monolingual model for hate speech classification of social media content in Slovenian language. The model was trained on 50,000 Twitter comments and tested on an independent test set of 10,000 Twitter comments. It is based on multilingual CroSloEngual BERT pre-trained language model.

Please cite:

Kralj Novak, P., Scantamburlo, T., Pelicon, A., Cinelli, M., Mozetič, I., & Zollo, F. (2022, July). Handling disagreement in hate speech modelling. In International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems (pp. 681-695). Cham: Springer International Publishing. https://link.springer.com/chapter/10.1007/978-3-031-08974-9_54

Tokenizer

During training the text was preprocessed using the original CroSloEngual BERT tokenizer. We suggest the same tokenizer is used for inference.

Model output

The model classifies each input into one of four distinct classes:

• 0 - acceptable
• 1 - inappropriate
• 2 - offensive
• 3 - violent