Trefferquote (Recall)
Definition
Recall measures how many truly relevant cases are found.
The proportion of truly relevant or positive cases that a system correctly identifies.
Simply explained
Of all truly defective parts, the proportion recognized by the system is counted.
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 100 truly defective parts, a model recognizes 92; the recall is 0.92.
Limits and pitfalls
- Recall alone does not show how many found hits are false. Precision, class definition, and decision threshold must be considered together.
Origin & history
Recall originates from information retrieval and statistical classification.
Use in mid-sized companies
- Possible application: In quality control, measure the proportion of all truly defective parts a model detects and which defects it misses.
Frequently asked questions
What does recall mean?
The proportion of actually relevant or positive cases that a system correctly finds. What mistakes should be avoided? Evaluating recall without precision, class definition, or decision threshold.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: