Urdu language news articles organized into 5 intent categories for veracity classification. It provides text data in the `news` column alongside `label` and `category` fields to support misinformation detection in a low-resource language.
Use Cases
- Train a binary classifier to identify misinformation in Urdu using the `news` and `label` columns.
- Perform multi-class intent classification on news articles using the `category` field.
- Benchmark Urdu language models on text classification tasks using the `news` string data.
Strengths
- Includes Urdu-script text strings in the `news` column.
- Provides a `label` field for binary classification of real and fake news.
- Contains a `category` column representing 5 distinct intents for the news content.