Microsoft DP-900

Unstructured data has no predefined schema and doesn’t naturally fit into rows and columns. Examples include PDFs, emails (as raw text), images, audio, video, and social media posts. While you can extract structure from it using AI/ML or parsing, the data itself isn’t stored in a consistent tabular format like the example provided.

Semi-structured data has some organization (keys/tags/metadata) but does not enforce a rigid, fixed schema across all records. Common examples are JSON, XML, YAML, and log files where fields may vary by event type. The given dataset shows a fixed set of columns for every row, which aligns with structured rather than semi-structured data.

A document typically refers to a self-contained data object (often JSON) stored in a document database (for example, Azure Cosmos DB for NoSQL). Although a single row could be represented as a document, the question’s tabular representation with consistent columns is characteristic of relational structured data, not the document data model emphasized in DP-900.

Structured data is organized into a predefined schema, commonly represented as tables with rows and columns. The example has consistent fields (ID, First name, Last name, Email) for each record, enabling SQL querying, constraints, and indexing. This is the classic definition of structured data used in relational databases and tabular datasets.