metadata

dataset_info:
  features:
    - name: url
      dtype: string
    - name: title
      dtype: string
    - name: text
      dtype: string
    - name: publish_date
      dtype: string
    - name: update_date
      dtype: string
    - name: scrape_date
      dtype: string
  splits:
    - name: acibadem
      num_bytes: 59871616
      num_examples: 6339
    - name: anadolusaglik
      num_bytes: 7711085
      num_examples: 1012
    - name: atlas
      num_bytes: 659835
      num_examples: 130
    - name: baskentistanbul
      num_bytes: 1632855
      num_examples: 394
    - name: bayindir
      num_bytes: 3985692
      num_examples: 690
    - name: florence
      num_bytes: 12815638
      num_examples: 1641
    - name: guven
      num_bytes: 4279987
      num_examples: 666
    - name: liv
      num_bytes: 19820891
      num_examples: 2836
    - name: medicalpark
      num_bytes: 3323029
      num_examples: 371
    - name: medicalpoint
      num_bytes: 2374940
      num_examples: 654
    - name: medicana
      num_bytes: 17090402
      num_examples: 2163
    - name: medipol
      num_bytes: 7876549
      num_examples: 1380
    - name: memorial
      num_bytes: 44864194
      num_examples: 5338
    - name: yeditepe
      num_bytes: 5203404
      num_examples: 998
  download_size: 89711227
  dataset_size: 191510117
configs:
  - config_name: default
    data_files:
      - split: acibadem
        path: data/acibadem-*
      - split: anadolusaglik
        path: data/anadolusaglik-*
      - split: atlas
        path: data/atlas-*
      - split: baskentistanbul
        path: data/baskentistanbul-*
      - split: bayindir
        path: data/bayindir-*
      - split: florence
        path: data/florence-*
      - split: guven
        path: data/guven-*
      - split: liv
        path: data/liv-*
      - split: medicalpark
        path: data/medicalpark-*
      - split: medicalpoint
        path: data/medicalpoint-*
      - split: medicana
        path: data/medicana-*
      - split: medipol
        path: data/medipol-*
      - split: memorial
        path: data/memorial-*
      - split: yeditepe
        path: data/yeditepe-*
license: cc-by-4.0
language:
  - tr
task_categories:
  - text-classification
  - text-generation
  - summarization
  - question-answering
tags:
  - medical
  - healthcare
  - turkish
  - hospital
  - articles
  - clinical-text
size_categories:
  - 10K<n<100K

🏥 Turkish Hospital Medical Articles Dataset

A comprehensive collection of Turkish-language medical articles from 14 major hospital and healthcare provider websites in Turkey. This dataset is designed for training and evaluating Turkish NLP models in the medical domain, including large language models (LLMs), health chatbots, medical text summarization, and clinical text classification.

📊 Dataset Overview

Total Articles: ~24,612 medical articles
Sources: 14 major Turkish hospitals and healthcare providers
Language: Turkish (tr)
Format: Parquet files (one per hospital)
License: CC BY 4.0
Total Size: ~95MB

🏥 Included Hospital Sources

Hospital	Dataset Split	Description
Acıbadem Hospital	`acibadem`	Leading private hospital network
Anadolu Sağlık	`anadolusaglik`	Comprehensive healthcare provider
Atlas Hospital	`atlas`	Specialized medical center
Başkent İstanbul	`baskentistanbul`	University hospital network
Bayındır Hospital	`bayindir`	Private healthcare institution
Florence Nightingale	`florence`	Specialized hospital network
Güven Hospital	`guven`	Private medical center
Liv Hospital	`liv`	International healthcare provider
Medical Park	`medicalpark`	Private hospital network
Medical Point	`medicalpoint`	Specialized medical facility
Medicana Hospitals	`medicana`	Large hospital network
Medipol University Hospital	`medipol`	University medical center
Memorial Hospital	`memorial`	Private healthcare network
Yeditepe University	`yeditepe`	University hospital

📋 Data Schema

Each article in the dataset contains the following fields:

Field	Type	Description
`url`	string	Original article URL
`title`	string	Article title
`text`	string	Full article content
`publish_date`	string	Original publication date
`update_date`	string	Last update date
`scrape_date`	string	Date when article was collected

🚀 Usage Examples

Load the entire dataset:

from datasets import load_dataset

# Load all hospital data
dataset = load_dataset("alibayram/turkish-hospital-medical-articles", token="your_token")

# Access specific hospital data
acibadem_data = dataset["acibadem"]
memorial_data = dataset["memorial"]

Load specific hospitals:

from datasets import load_dataset

# Load only specific hospitals
selected_hospitals = {
    "acibadem": "acibadem.parquet",
    "memorial": "memorial.parquet",
    "medicana": "medicana.parquet"
}

dataset = load_dataset("alibayram/turkish-hospital-medical-articles",
                      data_files=selected_hospitals,
                      token="your_token")

Access article examples:

# Get first article from Acıbadem
first_article = dataset["acibadem"][0]
print(f"Title: {first_article['title']}")
print(f"Content: {first_article['text'][:200]}...")
print(f"Published: {first_article['publish_date']}")

🎯 Use Cases

Medical NLP Applications:

Medical Text Classification: Categorize articles by medical specialty
Health Information Extraction: Extract symptoms, treatments, and medical procedures
Medical Question Answering: Train models to answer health-related questions
Medical Text Summarization: Create concise summaries of medical articles
Health Chatbot Training: Develop Turkish medical chatbots
Clinical Text Analysis: Analyze medical terminology and concepts

Research Applications:

Medical Language Modeling: Train Turkish medical LLMs
Healthcare Information Retrieval: Build medical search systems
Medical Knowledge Graph Construction: Extract medical entities and relationships
Cross-hospital Analysis: Compare medical content across different institutions

🔧 Data Processing

The dataset has been processed to:

Remove unnecessary columns (headings, source metadata)
Standardize date fields
Ensure consistent schema across all hospital sources
Maintain original article content integrity

📈 Dataset Statistics

Total Articles: 24,612
Average Article Length: ~500-2000 words
Date Range: Various (based on hospital website content)
Medical Topics Covered:
- General medicine
- Specialized treatments
- Health tips and advice
- Medical procedures
- Disease information
- Wellness and prevention

⚠️ Important Notes

Medical Disclaimer: This dataset contains medical information but should not be used for clinical decision-making
Language: All content is in Turkish
Source Attribution: Articles are from official hospital websites
Privacy: No personal patient information is included
Accuracy: Content accuracy depends on the original hospital websites

🤝 Contributing

This dataset is based on the original work by umutertugrul and has been processed and organized for improved usability.

📄 License

This dataset is licensed under the Creative Commons Attribution 4.0 International License.

🙏 Acknowledgments

Original data collection by umutertugrul
Hospital websites for providing public medical content
Hugging Face for hosting the dataset

For questions or issues, please open an issue on the dataset page.