Data Management & Integration Flashcards
7 cards from real GMI practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 Data Management & Integration flashcards as text
A printing facility collects L*a*b* data from multiple spectrophotometers of different models. Before merging datasets, what must be verified?
Answer: That all instruments are calibrated to the same illuminant and observer conditions
Measurement conditions (illuminant such as D50, observer angle such as 2°) must be identical across instruments to ensure the L*a*b* values are directly comparable.
Which ETL stage is responsible for converting density readings from a densitometer's native format into the standardized database schema?
Answer: Transform
The Transform stage converts, cleans, and maps source data into the target schema format before it is loaded into the destination database.
When a client requires real-time access to press measurement data, which integration architecture is most suitable?
Answer: Event-driven API with WebSocket streaming
An event-driven API with WebSocket streaming pushes measurement updates instantly as they occur, enabling true real-time client visibility.
In a GMI-compliant data management system, what is the purpose of a measurement audit trail?
Answer: To record who measured what, when, and with which instrument
An audit trail logs the operator, instrument, timestamp, and conditions for every measurement, supporting traceability and dispute resolution.
Which SQL statement is used to combine measurement records from two separate press run tables while eliminating duplicate rows?
Answer: UNION
UNION combines result sets from two SELECT queries and automatically removes duplicate rows, unlike UNION ALL which retains them.
A color management database stores ICC profile metadata alongside measurement data. Which field should serve as the foreign key linking measurements to their associated ICC profile?
Answer: Profile UUID or unique identifier
A UUID (universally unique identifier) guarantees referential integrity as a foreign key because it is immutable and globally unique, unlike descriptions or dates.
When validating imported spectral reflectance data, which check confirms that values are physically plausible?
Answer: Verifying that all reflectance values fall between 0.0 and 1.0 (or 0% and 100%)
Spectral reflectance is a ratio of reflected to incident light and must physically fall between 0 (no reflection) and 1 (perfect reflection), so out-of-range values signal corruption.