metadata
dataset_info:
features:
- name: tweet
dtype: string
- name: sentiment
dtype: int64
splits:
- name: train
num_bytes: 147321
num_examples: 948
- name: test
num_bytes: 36876
num_examples: 237
download_size: 100220
dataset_size: 184197
arxiv: arxiv.org/abs/2304.04703
task_categories:
- text-classification
language:
- ku
tags:
- Kurdish
- Central Kurdish
- Sorani
- ckb
size_categories:
- 1K<n<10K
Central Kurdish Sentiment Dataset
This Central Kurdish dataset is compiled from the transfer learning project for sentiment analysis. It contains 1185 tweets labeled with positive, negative, and neutral sentiments.