All projects
Turkish Legal RAG Benchmark
Controlled evaluation of four RAG architectures across 422 legal documents and 1,030 questions (TUBITAK 2209-A).
2025 TUBITAK 2209-A Researcher
- Python
- FAISS
- BGE-M3
- FastAPI
- React
- SSE
ARCHITECTURE NOTE
Benchmarked four retrieval configurations over 5,979 chunks using paired statistical significance tests; reached 85.3% retrieval recall, 0.472 answer F1, and 0.790 composite RAGAS score with cross-encoder reranking, backed by a source-citing SSE Q&A prototype.
Overview
Controlled retrieval-augmented generation benchmark funded by TUBITAK 2209-A, paired with a full-stack legal Q&A prototype featuring hybrid FAISS + lexical retrieval, BGE-M3 embeddings, SSE token streaming, and inline statutory citations.