Research and development at Lanka Data Foundation.
Legislation Project Documentation
The Legislation project provides a comprehensive platform for researching, analyzing, and navigating Sri Lankan legislative documents. It includes semantic search, AI-powered analysis, and lineage tracking.
This project explores OCR capabilities using DeepSeek models on the Hugging Face platform, providing scripts and notebooks for extracting text from scanned documents and evaluating performance.
Gazette Analysis Documentation
Tools for extracting and processing Sri Lankan government gazette data. This project includes:
- GztExtractor: LLM-based extraction of ministry structures, amendments, and personnel appointments from gazette PDFs
- GztProcessor: A versioning system to track structural changes in government organizations over time, with FastAPI backend and React frontend
- Gazette Tracer: Tools for tracing and visualizing gazette relationships using Neo4j graph database, with Flask API and React UI