An AI-powered q & a agent system focused on Islamic knowledge, leveraging scraped content from authentic sources and OpenAI's NLP capabilities. The system provides context-aware answers about Islamic practices, beliefs, and traditions while maintaining source fidelity.
- Q&A Interface: Natural language interface for asking Islamic-related questions
- Context-Aware Answers: AI responses grounded in verified Islamic content
- Knowledge Base:
- Pre-loaded sample data covering fasting (Oruç), Ramadan, and prayer (Dua)
- Web crawler integration for islamveihsan.com
- Search Index: Fast content retrieval system using Whoosh search engine
- Validation System: Ensures answers stay within provided Islamic context
- Batch Processing: Automated setup script for data/index preparation
The system operates in 3 phases:
- Data Preparation: Creates sample data & crawls a given websites
- Index Building: Creates searchable knowledge base
- Q&A Interface: Interactive Islamic knowledge assistant
# Start the system (Windows) - Optional
.\run_project.bat
# Run project (Windows)
Python main.py
# Or run manually:
python create_sample_data.py
python indexer/build_index.py
python main.pycreate_sample_data.py: Creates initial dataset with Islamic rulingsindexer/: Search index builder using Whooshcrawler/: Website spider for islamveihsan.commain.py: Core AI response generationconfig/: Contains API and website configurations
Ibrahim: @iBz-04
