Datasets:
Dataset Viewer
The dataset viewer is not available for this dataset.
Unexpected token '<', "<html>
<h"... is not valid JSON
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Dataset Summary
This is a diverse Turkish corpus consisting of 120,000 rows, categorized into 5 major domains.
Dataset Schema & Labels
| Category | Labels | Description |
|---|---|---|
| Conversation | 0 | Daily dialogues and conversational language |
| Review | 1 | User reviews and sentiment-related data |
| Wiki | 2 | Encyclopedic and formal informational texts |
| News | 3 | News articles and journalistic language |
| STEM | 4 | Academic texts (Science, Tech, Engineering, Math, Medicine) |
Sources
The dataset is curated from the following sources:
- Conversation:
TFLai/Turkish-Dialog-Dataset - Review:
winvoker/turkish-sentiment-analysis-dataset - Wiki:
wikimedia/wikipedia(Turkish) - News:
batubayk/TR-News - STEM: *
hazal/Turkish-Biomedical-corpus-trMhazal/electronic-radiology-phd-thesis-trRysdede/khanacademy-turkish
- Downloads last month
- 14