1.4 Ranking Metrics #4
Description
Goal
Implement evaluation metrics for search systems.
Learn
- Precision@k, Recall@k
- Mean Reciprocal Rank (MRR)
- NDCG (Normalized Discounted Cumulative Gain)
- Building relevance judgment sets
Deliverable
- Evaluation harness that scores any ranker
- Annotated relevance judgments for test queries
- Comparison of TF-IDF, BM25, semantic on metrics
Proof Point
Can explain what NDCG=0.7 means and why position matters.
Directory
search-fundamentals/04-ranking-metrics/