Hate Speech Classifier Evaluation Project

Overview

This project makes it easy to annoatated and teste a dataset using different hate speech classifiers for different languages, including German, English, and Multilingual. The classifiers are provided through Hugging Face and include models from various sources. The project takes a simple CSV file with a "text" column as input and outputs separate CSV files for each classifier. Classifier as set to be bianry hate no hate at the moment.

Classifiers and URLs

1. Google Jigsaw Classifier - Note: An API key is required for this classifier.

2. IMSyPP Hate Speech Classifier (English)

3. Hatexplain BERT Classifier

4. RoBERTa Hate Speech Classifier (Facebook)

5. XLM-Roberta Large Hate Speech Classifier (Multilingual)

6. Multilingual Hate Speech Classifier (Robacofi)

Input and Output

• Input: The project accepts a simple CSV file with a column named "text" containing the text data to be classified.

• Output: Separate CSV files are generated for each classifier, containing the original text along with the classification results.

Contribution

Contributions are welcome!

License

tbd