GCP Generative AI Leader Practice Question 11: Structured Unstructured Data

Google Cloud defines unstructured data as information that doesn't follow a predefined format, model, or structure; it doesn't fit neatly into rows and columns, and it shows up as text files, emails, videos, images, audio recordings, and social media posts. Structured data has a strict, predefined schema, lives in relational databases, and is queried with SQL; financial records and inventory are the classic examples. That puts the call recordings, scanned documents, videos, and complaint emails in the unstructured group and the core banking tables in the structured group (option C). The distinction matters to a leader because unstructured data is the bulk of what most enterprises hold and is exactly the material gen AI can finally read at scale.

- A is wrong because a file extension describes the container, not the data model. A .mp4 or .eml file still has no predefined schema you can query with SQL. - B is wrong because the classification depends on whether the data has a predefined model, not on why it was originally collected. The banking tables are plainly structured. - D inverts the two definitions. Rows changing frequently is a matter of freshness, not structure; a transactions table with a fixed schema is structured no matter how often it is updated.

Reference: What is unstructured data?