DUAA Lesson Content Generator

An AI-powered question generation system that extracts content from PDF files and generates Open edX-compatible questions using AWS Bedrock Claude models.

Features

• PDF Content Extraction: Automatically extracts text from PDF course materials
• AI Question Generation: Uses AWS Bedrock Claude 3.5 Sonnet for intelligent question creation
• Open edX Integration: Publishes generated questions directly to Open edX content libraries
• Multiple Question Types: Supports single-select, multi-select, dropdown, numerical input, and text input questions
• Flexible Configuration: Environment-based configuration for different courses and libraries

Architecture

Question Generation Pipeline

PDF Content → Text Extraction → AWS Bedrock Claude → Question Generation → Open edX Publishing

1. Content Ingestion: PDF files are processed and text is extracted
2. AI Processing: AWS Bedrock Claude 3.5 Sonnet generates diverse questions based on content
3. Question Formatting: Questions are converted to Open edX OLX format
4. Library Publishing: Questions are published to specified Open edX content libraries

AWS Bedrock Integration

This system uses AWS Bedrock's Claude 3.5 Sonnet model (anthropic.claude-3-5-sonnet-20240620-v1:0) for question generation. The AI model:

• Analyzes PDF content and learning objectives
• Generates diverse question types appropriate for different cognitive levels
• Creates questions with proper explanations and metadata
• Ensures questions align with Bloom's taxonomy levels

Supported Question Types

• Single Select: Multiple choice with one correct answer
• Multi-select: Multiple choice with multiple correct answers
• Dropdown: Selection from dropdown menu
• Numerical Input: Number-based answers with tolerance
• Text Input: Short text responses

Installation

1. Clone the repository:

git clone <repository-url>
cd duaa-lesson-content-generator

2. Install dependencies:

pip install -r requirements.txt

3. Configure AWS credentials:

# Set AWS credentials for Bedrock access
export AWS_ACCESS_KEY_ID=your_access_key
export AWS_SECRET_ACCESS_KEY=your_secret_key
export AWS_DEFAULT_REGION=us-east-1

Configuration

Environment Variables

# AWS Bedrock Configuration
export AWS_ACCESS_KEY_ID=your_access_key
export AWS_SECRET_ACCESS_KEY=your_secret_key
export BEDROCK_MODEL_ID=anthropic.claude-3-5-sonnet-20240620-v1:0

# Open edX Configuration
export OPENEDX_API_URL=http://studio.local.openedx.io:8001
export OPENEDX_LIBRARY_ID=lib:myorg:kg1_english_2025
export CSRF_TOKEN=your_csrf_token
export STUDIO_SESSION_ID=your_session_id
export JWT_HEADER_PAYLOAD=your_jwt_payload

# Content Configuration
export COURSE_CONTENT_DIR=/path/to/course/content

Alternative: OpenAI Fallback

If AWS Bedrock is not available, the system can fallback to OpenAI:

export OPENAI_API_KEY=your_openai_key

Usage

Basic Usage

from src.main import generate_questions_for_openedx

# Generate questions from PDF content
result = generate_questions_for_openedx(
    course_content_dir="./Course Content",
    library_id="lib:myorg:kg1_english_2025",
    topic="KG1 English Course",
    num_questions=10,
    aws_region='us-east-1'
)

if result['success']:
    print(f"Generated {result['questions_generated']} questions")
    print(f"Published {result['questions_published']} questions")

Command Line Usage

python src/main.py

Creating Day-Level Collections

The system supports creating organized question collections for different course days:

from src.utils.openedx_hierarchy import OpenEdXHierarchyManager

# Create day-level collections
manager = OpenEdXHierarchyManager(library_id, api_url, cookies)

for day in range(1, 11):  # Day 1-10
    collection = manager.create_collection(
        f"Day {day} Questions",
        f"Questions for Day {day} content"
    )
    
    # Generate and publish day-specific questions
    questions = generate_questions_with_bedrock(
        day_content, f"Day {day} Content", prompt, num_questions=5
    )
    publish_to_openedx(library_id, questions, api_url, cookies, collection['key'], day)

Project Structure

duaa-lesson-content-generator/
├── src/
│   ├── main.py                    # Main entry point
│   ├── openai_utils/
│   │   └── prompts.py            # AI prompts for question generation
│   └── utils/
│       ├── question_generator.py  # Core question generation logic
│       └── openedx_hierarchy.py  # Open edX collection management
├── requirements.txt              # Python dependencies
└── README.md                    # This file

Dependencies

• boto3: AWS SDK for Bedrock access
• openai: OpenAI API client (fallback)
• PyPDF2: PDF text extraction
• requests: HTTP client for Open edX API
• python-dotenv: Environment variable management

Open edX Integration

The system integrates with Open edX through the tutor-contentlibrary-api plugin:

1. Authentication: Uses CSRF tokens and session cookies
2. Question Publishing: Creates problem blocks in content libraries
3. OLX Format: Generates Open edX-compatible XML question format
4. Collection Management: Organizes questions into day-level collections

Error Handling

The system includes comprehensive error handling for:

• PDF extraction failures
• AWS Bedrock API errors
• Open edX authentication issues
• Network connectivity problems
• Invalid question formats

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

For issues and questions:

1. Check the error logs for detailed error messages
2. Verify AWS credentials and permissions
3. Ensure Open edX API access is properly configured
4. Review environment variable settings

Changelog

Version 1.0.0

• Initial release with AWS Bedrock integration
• Open edX question publishing
• PDF content extraction
• Multiple question type support
• Day-level collection organization