3.2Standard
Initial search returns many candidates but poor top evidence.
Measure the effect of adding a reranker.
Choose two evaluations.
× Incorrect
Compare candidate recall, top-k evidence accuracy and diversity, latency, cost, authorization filtering, and freshness before and after reranking.
Detailed explanation
IncorrectDecide from one attractive top result
Input diversity and missed candidates are hidden.
IncorrectUse reranking as an authorization replacement
Ordering does not grant access.
CorrectCompare top-k evidence accuracy and diversity while preserving candidate recall
Top quality and recall are both measured.
CorrectMeasure latency, token cost, authorization filtering, and stale-document rank
Operational and governance effects are included.
Check official documentation
Compare the following points with the official documentation.
Amazon Bedrock Knowledge Bases公式の再ランキング、検索評価、AIF-C01 Domain 3.2を確認する。
Official references
What to confirm
再ランキングの品質向上と費用・権限の副作用を評価できる。
Key points
- Recall
- Top-k
- Authorization
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Compare the definition with the AIF-C01 objective and the service documentation before applying it.
- Keep the concept separate from a particular answer choice or implementation step.
Foundation review
Reranking
Reordering initial retrieval candidates with additional features or a second model.
Recall
The share of all relevant items that a retrieval process returns.
Question IDAWS-AIF-282
EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
Last technical review2026-08-20
Report an error or rights issue