GCP Generative AI Leader Practice Question 23: Data Completeness Quality
Google Cloud Generative AI Leader · Test 3 · question 3 of 10
A retailer wants to ground a gen AI shopping assistant in its product catalog. During review, the team finds that many product records are missing the size, material, or return-policy fields, even though the records that do have them are correctly spelled and formatted.
Which data quality problem does this describe, and why does it matter for the assistant?
Show answer and explanation
Answer:
The scenario is about missing fields, not wrong or contradictory ones: it is a completeness problem. Completeness measures whether a dataset contains all the information it is expected to have. When required fields such as size or return policy are absent, a gen AI assistant grounded on that data either has to say it does not know or, worse, may hallucinate a plausible-sounding answer for the missing detail, which erodes trust.
Option B describes inconsistency, where the same fact conflicts across records. Here the existing fields are described as correctly spelled and formatted, so the problem is absence, not conflict.
Option C describes staleness or freshness, a separate dimension about data going out of date over time. Nothing in the scenario says the data was once complete and has since aged.
Option D describes duplication, where the same entity is recorded more than once. The scenario describes single records with empty fields, not repeated records.
Reference: Auto data quality overview