Microsoft Certified: Azure Data Fundamentals (DP-900) Practice Test

Semi-structured data is characterized by a lack of a fixed schema, meaning it does not conform to the strict structure of traditional databases but still contains some organizational properties that make it easier to analyze. JSON (JavaScript Object Notation) is a prime example of semi-structured data because it uses a flexible format with key-value pairs, allowing for hierarchical organization.

In JSON, data objects can vary in structure, meaning that different instances of a data entry can have different attributes or nested structures, while still maintaining the overall conventions of JSON syntax. This adaptability makes JSON ideal for representing complex data relationships and is commonly used in web applications and APIs.

Other formats, such as SQL databases, possess a rigid schema that enforces structure and type for each piece of data, making them structured. CSV files, while they do allow for a simple representation of tabular data, are generally considered structured as well since they follow a clear and consistent format of rows and columns. Binary data lacks a defined structure altogether and is not human-readable, placing it outside the scope of semi-structured data. Thus, JSON stands out as the correct answer for this question.