zabiullah/pdf-books-collection

📚 PDF Books Collection with Metadata

Dataset Description

This dataset contains 1043 PDF textbooks and educational materials (13.56 GB) covering various technical and philosophical topics.

🎯 Topics Covered

• Computer Science: Programming, Algorithms, Operating Systems, Software Engineering
• Mathematics: Calculus, Algebra, Geometry, Statistics, Mathematical Thinking
• Artificial Intelligence: Machine Learning, Deep Learning, Neural Networks, LLMs, ChatGPT
• Data Science: Analytics, Visualization, Data Mining
• Physics: Quantum Mechanics, Theoretical Physics, Universe & Cosmology
• Buddhism & Philosophy: Zen, Meditation, Buddhist Thought, Critical Thinking
• Networking: IoT, Cloud Computing, 5G, Infrastructure
• Security: Cryptography, Forensics, Cybersecurity

📁 Dataset Structure

Each row in the dataset contains:

{
    'file_name': str,      # Original PDF filename
    'title': str,          # Book title
    'author': str,         # Author(s) - may be empty
    'subject': str,        # Subject/topic - may be empty
    'language': str,       # Language code - may be empty
    'category': str,       # Auto-categorized topic (if available)
    'pdf': bytes          # The actual PDF file content
}

💻 Usage

Load the entire dataset:

from datasets import load_dataset

# Load dataset
dataset = load_dataset("YOUR_USERNAME/pdf-books-collection")

# Access a book
book = dataset['train'][0]
print(f"Title: {book['title']}")
print(f"Author: {book['author']}")

# Save PDF to disk
with open(f"{book['file_name']}", 'wb') as f:
    f.write(book['pdf'])

Stream for large datasets:

# Stream without downloading everything
dataset = load_dataset("YOUR_USERNAME/pdf-books-collection", streaming=True)

# Iterate through books
for book in dataset['train']:
    print(book['title'])
    # Process or save as needed

Filter and search:

# Load as pandas for easy filtering
df = dataset['train'].to_pandas()

# Find machine learning books
ml_books = df[df['title'].str.contains('Machine Learning', case=False)]

# Find books by author
author_books = df[df['author'].str.contains('Russell', case=False)]

# Filter by category
ai_books = df[df['category'] == 'AI & Machine Learning']

📊 Statistics

• Total Books: 1043
• Total Size: 13.56 GB
• Format: PDF
• Languages: Primarily English
• Metadata Completeness:
  • Titles: 100% (extracted or from filename)
  • Authors: ~65%
  • Subjects: ~40%
  • Languages: ~20%

🎓 Educational Use Cases

1. Building a Personal Library System: Create searchable digital library
2. Research & Study: Quick access to textbooks across multiple domains
3. Machine Learning Projects:
   • Document classification
   • OCR and text extraction
   • Semantic search implementation
   • RAG (Retrieval Augmented Generation) systems
4. Educational AI: Training material for educational chatbots
5. Bibliography Analysis: Study trends in technical literature

⚖️ Legal & Ethical Considerations

IMPORTANT: This dataset contains copyrighted educational materials. Users should:

• ✅ Use for personal study and research
• ✅ Verify copyright status before commercial use
• ✅ Respect author and publisher rights
• ❌ Not redistribute commercially without permission
• ❌ Not use for piracy or copyright infringement

This collection is shared for educational and research purposes under fair use principles. If you are a copyright holder and want content removed, please contact the dataset maintainer.

🔧 Technical Details

• Extraction Method: PyPDF2 for metadata extraction
• Storage Format: Parquet with binary PDF column
• Hosting: Hugging Face Datasets (Community Tier)
• Access: Public dataset, free to download

📝 Citation

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

@dataset{pdf_books_collection_2025,
  author = {Your Name},
  title = {PDF Books Collection with Metadata},
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/YOUR_USERNAME/pdf-books-collection}}
}

🤝 Contributing

Found an error in metadata? Want to add more books? Please open an issue or pull request!

📧 Contact

For questions, issues, or takedown requests, please open an issue in this repository.

---

Note: This is a community-contributed dataset. The maintainer is not responsible for the copyright status of individual files. Users are responsible for ensuring their use complies with applicable laws and regulations.