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-trM
    • hazal/electronic-radiology-phd-thesis-trR
    • ysdede/khanacademy-turkish
Downloads last month
14