What does data lineage refer to in CPMAI data governance?

What does data lineage refer to in CPMAI data governance?

• A. The performance metrics of the model.
• B. The software dependency graph.
• C. The user access control settings.
• D. The origin and transformations of data through the pipeline, enabling traceability.

Answer: (D)

In data governance, data lineage is about where data comes from and how it moves and changes as it flows through the pipeline. It tracks the data’s origin, every transformation it undergoes (such as cleaning, normalization, feature extraction), and where it goes next, providing end-to-end traceability. This is crucial in CPMAI because AI systems rely on data for training and inference; knowing the provenance and transformations helps explain model behavior, audit data quality, reproduce experiments, and assess risk or regulatory compliance. This isn’t about how well the model performs, nor about the software components that data flows through, nor about who can access the data. Those serve other governance purposes; data lineage specifically maps the data’s origin and its journey through the pipeline to enable traceability.

In data governance, data lineage is about where data comes from and how it moves and changes as it flows through the pipeline. It tracks the data’s origin, every transformation it undergoes (such as cleaning, normalization, feature extraction), and where it goes next, providing end-to-end traceability. This is crucial in CPMAI because AI systems rely on data for training and inference; knowing the provenance and transformations helps explain model behavior, audit data quality, reproduce experiments, and assess risk or regulatory compliance.

This isn’t about how well the model performs, nor about the software components that data flows through, nor about who can access the data. Those serve other governance purposes; data lineage specifically maps the data’s origin and its journey through the pipeline to enable traceability.