Context recall
A RAG retrieval metric: how much of the information needed for the correct answer was actually retrieved.
The reference answer is split into claims and each is checked against the retrieved context; the score is the share that can be found there. It needs ground truth, so it is run on a prepared test set. A low score means the answer was lost before the model saw anything: the document is missing from the index, was split badly or was not ranked high enough.
Related terms
- Context precision: A RAG retrieval metric: whether the passages relevant to the question are ranked at the top of what was retrieved.
- Recall (AI metric): The share of all correct items that the system actually found: of everything it should have caught, how much it did.
- Ground truth: The known correct answers or labels that a model’s output is compared against in an evaluation.
- Chunking (RAG): Splitting documents into smaller passages before indexing them for search, so retrieval returns the relevant part instead of a whole file.
- Ragas: An open-source Python library for evaluating RAG systems with metrics such as faithfulness, answer relevancy, context precision and context recall.
