The origin of each portion: To satisfy data lineage and provenance requirements, what must the provenance
Provenance for a mixed dataset must trace the origin of every part, including how synthetic data was generated and how sources were combined.
The question
A team assembles a training set by combining licensed third-party data with synthetic records it generated to fill gaps. To satisfy data lineage and provenance requirements, what must the provenance record capture for this mixed dataset?
Preparing for AIGP? Take the free 5-min readiness quiz →
- The origin of each portion, including third-party sources and the method used to generate the synthetic data, plus how the parts were combined. ✓Provenance must trace where each part came from, including how synthetic records were generated and how the sources were merged, for a mixed dataset.
- Only the license terms of the third-party portion, since the synthetic records were created internally and therefore carry no meaningful provenance to record.Synthetic data has provenance too, namely how and from what it was generated; omitting it leaves the dataset's origin partly untraceable.
- Only the aggregate row count and column schema of the merged set, since documenting the final structure is what allows the data to be reused reliably later.Schema and counts describe structure, not origin, so they do not satisfy lineage and provenance for the combined dataset.
- Only the model's performance on the merged data, since demonstrating that the combined set produces good results is what validates its provenance in practice.Performance validates utility, not origin; provenance is about tracing where the data came from, which results cannot show.
The trap
Assuming internally generated synthetic data has no provenance worth documenting. How to remember it
Provenance for a mixed dataset must trace the origin of every part, including how synthetic data was generated and how sources were combined.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding How to Govern AI Development questions
- The data is not fit-for-purpose because its population is: Applying data-governance requirements, which →
- Security testing that probes the pipeline and model: Within the training-and-testing plan, which test type →
- Investigate the disparity's source in data or design: Under III.B.4, what is the most appropriate way to →
- All 426 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.
