PropXplain is a multilingual dataset for explainable propaganda detection in Arabic and English text. It provides both propaganda classification labels and natural language explanations, enabling the development of interpretable propaganda detection systems.
How to Use
To load the dataset, you need to specify the language configuration (either "arabic" or "english").
from… See the full description on the dataset page: https://huggingface.co/datasets/QCRI/PropXplain.
