1,800 annotated tweets categorized into positive and negative sentiment classes. The collection features text written in both Modern Standard Arabic (MSA) and the specific Jordanian dialect.
Use Cases
- Train a sentiment classifier to distinguish between positive and negative sentiments in Jordanian dialect text
- Evaluate the performance of Arabic NLP models on regional dialectal variations compared to Modern Standard Arabic
- Perform linguistic analysis on the differences between MSA and Jordanian dialect in social media contexts
Strengths
- 1,800 manually annotated tweets
- Binary sentiment labels consisting of 'positive' and 'negative'
- Dual-language coverage including Modern Standard Arabic (MSA) and Jordanian dialect