winvoker/milyoner-soru-cevap

Turkish Quiz Show Dataset (Milyoner)

Dataset Description

This dataset contains questions and answers from the Turkish quiz show "Milyoner" (Turkish version of "Who Wants to Be a Millionaire?"). The dataset includes multiple-choice questions across various categories, contestant responses, joker usage, and game outcomes.

Dataset Structure

The dataset contains 3,228 questions from 68 episodes with 439 anonymized contestants.

Columns

Column	Type	Description
video_id	string	Unique identifier for the episode/video
contestant	string	Anonymized contestant identifier (e.g., "Contestant_001")
question	string	The quiz question text in Turkish
options	string	Multiple choice options as a list (A, B, C, D)
correct_answer	string	The correct answer option (A, B, C, or D)
contestant_answer	string	The contestant's chosen answer (A, B, C, D, or empty for no answer)
category	string	Question category (e.g., "Genel Kültür", "Edebiyat", "Müzik")
level	integer	Difficulty level (1-15, where 1 is easiest)
amount	float	Prize amount for the question in Turkish Lira
joker_used	string	Type of lifeline used ("yok", "seyirci", "telefon", "yarı_yarıya", "çift_cevap")
is_correct	boolean	Whether the contestant answered correctly
eliminated	boolean	Whether the contestant was eliminated after this question

Categories

The dataset includes questions from 24 different categories:

• Genel Kültür (General Knowledge) - Most common category
• Müzik (Music)
• Edebiyat (Literature)
• Tarih (History)
• Bilim (Science)
• Matematik (Mathematics)
• Coğrafya (Geography)
• Spor (Sports)
• Sanat (Arts)
• And 15 more specialized categories

Lifelines (Jokers)

• yok: No lifeline used
• seyirci: Ask the audience
• telefon: Phone a friend
• yarı_yarıya: 50:50 (eliminate two wrong answers)
• çift_cevap: Double answer (can give two answers)

Data Statistics

• Total Questions: 3,228
• Total Episodes: 68
• Unique Contestants: 439 (anonymized)
• Question Levels: 1-13 (difficulty levels)
• Categories: 24 different subject areas
• Languages: Turkish

Answer Distribution

• Correct Answers: ~75% of questions answered correctly
• Incorrect/No Answer: ~25% of questions
• Elimination Rate: Varies by difficulty level

Usage Examples

Loading the Dataset

import pandas as pd

# Load the dataset
df = pd.read_csv('milyoner_data_anonymized.csv')

# Basic statistics
print(f"Total questions: {len(df)}")
print(f"Unique contestants: {df['contestant'].nunique()}")
print(f"Categories: {df['category'].nunique()}")

Analyzing Question Difficulty

# Success rate by level
success_by_level = df.groupby('level')['is_correct'].mean()
print(success_by_level)

# Questions per category
category_counts = df['category'].value_counts()
print(category_counts)

Analyzing Lifeline Usage

# Lifeline usage frequency
joker_usage = df['joker_used'].value_counts()
print(joker_usage)

# Success rate when using lifelines
joker_success = df.groupby('joker_used')['is_correct'].mean()
print(joker_success)

Potential Use Cases

1. Question Answering Systems: Train models to answer Turkish quiz questions
2. Difficulty Assessment: Analyze question difficulty based on success rates
3. Educational Research: Study learning patterns and knowledge domains
4. Game Strategy Analysis: Examine optimal lifeline usage strategies
5. Natural Language Processing: Turkish text classification and analysis
6. Cultural Knowledge Mining: Extract cultural and educational insights

Data Quality Notes

• Questions are in Turkish and may contain some OCR/transcription errors
• Contestant names have been anonymized for privacy
• Some questions may have incomplete answer options
• Prize amounts may be in different formats due to show evolution

Privacy and Ethics

• All contestant names have been anonymized
• No personal information about contestants is included
• Data is derived from publicly broadcast television content
• Suitable for research and educational purposes

Citation

If you use this dataset in your research, please cite:

Turkish Quiz Show Dataset (Milyoner)
Source: Turkish Television Quiz Show "Milyoner"
Dataset Version: 2025
Total Questions: 3,228
Episodes: 68 (anonymized)

License

This dataset is provided for research and educational purposes. The original content is from publicly broadcast television shows.

File Information

• Filename: milyoner_data_anonymized.csv
• Size: ~2.5MB
• Encoding: UTF-8
• Format: CSV with comma separators

Contact

For questions about this dataset or to report issues, please open an issue in the repository. Or you can contact me via LinkedIn https://www.linkedin.com/in/batuhanayhan/

---

This dataset represents a valuable resource for Turkish natural language processing, educational research, and game show analysis.