About
This workflow applies Benford's Law First-Digit Analysis to an anonymized General Ledger (GL) extracted from a real accounting environment. The dataset is available on Kaggle:
🔗 Anonymized General Ledger Dataset
The focus is practical, not theoretical — specifically, analyzing multiple account code sub-groups independently, which is what real audit work demands. KNIME's Group Loop drives the per-segment iteration automatically.
Detection Logic
A two-level flagging approach is applied at each Account Code segment:
Dataset Level — MAD (Mean Absolute Deviation) classifies each segment against Nigrini's conformity thresholds
Digit Level — Gate 1 (Absolute Deviation) × Gate 2 (Z-Score); a digit is flagged only when both gates fire, keeping findings defensible
Notable Techniques
Empty Table Switch — handles zero-anomaly iterations cleanly
Email Connector with Attachment — delivers the anomaly report automatically, making this usable in a recurring audit cycle
