Keyboard shortcuts

Press or to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Use Cases

Real-world applications for SyntheticData.

Overview

Use CaseDescription
Fraud Detection MLTrain supervised fraud models
Audit AnalyticsTest audit procedures
SOX ComplianceTest control monitoring
Process MiningGenerate OCEL 2.0 event logs
ERP Load TestingLoad and stress testing

Use Case Summary

Use CaseKey FeaturesOutput Focus
Fraud DetectionAnomaly injection, graph exportLabels, graphs
Audit AnalyticsFull document flows, controlsTransactions, controls
SOX ComplianceSoD rules, approval workflowsControls, violations
Process MiningOCEL 2.0 exportEvent logs
ERP TestingHigh volume, realistic patternsRaw transactions

Quick Configuration

Fraud Detection

anomaly_injection:
  enabled: true
  total_rate: 0.02
  generate_labels: true

graph_export:
  enabled: true
  formats:
    - pytorch_geometric

Audit Analytics

document_flows:
  p2p:
    enabled: true
  o2c:
    enabled: true

internal_controls:
  enabled: true

SOX Compliance

internal_controls:
  enabled: true
  sod_rules: [...]

approval:
  enabled: true

Process Mining

document_flows:
  p2p:
    enabled: true
  o2c:
    enabled: true

# Use datasynth-ocpm for OCEL 2.0 export

ERP Testing

transactions:
  target_count: 1000000

output:
  format: csv

Selecting a Use Case

Choose Fraud Detection if:

  • Training ML/AI models
  • Building anomaly detection systems
  • Need labeled datasets

Choose Audit Analytics if:

  • Testing audit software
  • Validating analytical procedures
  • Need complete document trails

Choose SOX Compliance if:

  • Testing control monitoring systems
  • Validating SoD enforcement
  • Need control test data

Choose Process Mining if:

  • Using PM4Py, Celonis, or similar tools
  • Need OCEL 2.0 compliant logs
  • Analyzing business processes

Choose ERP Testing if:

  • Load testing financial systems
  • Performance benchmarking
  • Need high-volume realistic data

Combining Use Cases

Use cases can be combined:

# Fraud detection + audit analytics
anomaly_injection:
  enabled: true
  total_rate: 0.02
  generate_labels: true

document_flows:
  p2p:
    enabled: true
  o2c:
    enabled: true

internal_controls:
  enabled: true

graph_export:
  enabled: true

See Also