Skip to content

ISO/IEC 5259-4:2024 Artificial intelligence - Data quality for analytics and machine learning - Part 4: Data quality process framework


Summary

This standard establishes a common process framework for ensuring data quality in analytics and machine learning (ML). It covers training and evaluation data workflows including acquisition, preparation, labelling, evaluation and decommissioning for supervised, unsupervised, semi-supervised and reinforcement learning. It has gained signnificant global traction, becoming a European standard in 2025.


Key Takeaways

  • Provides a structured process model that helps organizations organize data-quality activities across ML and analytics workflows.
  • Establishes a consistent framework that organizations can adapt to different contexts.
  • Supports repeatability by outlining how processes should be sequenced and coordinated throughout data acquisition, preparation, and use.
  • Reinforces integration with the other parts of ISO/IEC 5259 by defining how process guidance complements measures (Part 2) and management requirements (Part 3).

Additional Sources


Tags

data, analytics, machine-learning, ml