Recall at k
Definition
Recall@k measures found relevant items within the top results.
The proportion of relevant items contained within the first k returned results.
Simply explained
One checks how many of the total matching documents appear within the first specified search ranks.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
Of four relevant documents, three are among the top ten hits; Recall@10 is 0.75.
Limits and pitfalls
- Choosing k without usage context, leaving relevance judgments incomplete, or confusing Recall@k with Precision@k.
Origin & history
Recall@k has become a standard metric with rank-based search and recommendation systems.
Use in mid-sized companies
- Possible application: Of four relevant documents, three are among the top ten hits; Recall@10 is 0.75.
Frequently asked questions
What does recall within the first k results mean?
The proportion of relevant items contained within the first k returned results. What mistakes should be avoided? Choosing k without usage context, leaving relevance judgments incomplete, or confusing Recall@k with Precision@k.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: