Chroma Vector Database for Master Thesis: Retrieval Optimization for RAG-based Chatbots

This dataset contains the Chroma vector database used in the Master's Thesis "Optimizing Retrieval Augmented Generation Chatbots: A Comparative Analysis". The database was built from publicly available directive documents and serves as the knowledge base for evaluating multiple retrieval approaches, including dense retrieval, parent-child hierarchical retrieval, and sparse retrieval with BM25. The data has been preprocessed and vectorized for use with Retrieval-Augmented Generation (RAG) chatbots.