Fig. 3From: Cleaning by clustering: methodology for addressing data quality issues in biomedical metadataRunning example showing the different variants of the key “age”Back to article page