Robust04 Retrieval: BM25 to Neural Reranking Pipeline

This document provides an overview of the TREC Robust04 dataset, details its indexing and setup procedures using Pyserini, and explores evaluation tools like MAP metrics. It analyzes retrieval model performance, including BM25, RM3 tuning, and neural reranking, highlighting their scores and improvements. Final results and future directions are discussed.