Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
json
Sub-tasks:
multi-label-classification
Languages:
Arabic
Size:
1K - 10K