schema

A schema is an abstract, formal blueprint that defines the structure, organization, or constraints of a collection of data or knowledge. It describes what kinds of data exist, how they relate, and how they can be validated. In addition to data schemas, the term is used in cognitive science to refer to mental structures that organize knowledge about concepts, events, or plans.

In databases, a schema specifies the organization of data in a database system. It defines tables, columns, data types, constraints such as primary keys and foreign keys, and relationships between tables. A schema provides a blueprint for data integrity and query planning, while an instance is the actual data stored in the schema. Schemas evolve through versioning, with considerations for backward compatibility.

In data formats and modeling, schemas constrain structure and content. XML and JSON schemas describe allowable elements, attributes, and data types; DTDs and XML Schema languages enforce validity. In data warehousing, schema design patterns such as star schemas and snowflake schemas organize dimensions and facts to support efficient querying.

Schemas are used for interoperability and validation, enabling tools to validate data, generate documentation, and improve search and automation. They are also central to semantic web schemas (for example, schema.org) which annotate web content for machine understanding.