A **data model** is a formal representation of the structure, relationships, and constraints governing data within a system or domain. Data models serve as blueprints that define how data is organized, how different data elements relate to one another, and what rules govern its integrity and use. They are foundational to [[Database design|database design]], [[Software engineering|software engineering]], [[Data architecture|data architecture]], and [[Information science|information science]]. Data models are conventionally described at three levels of abstraction. A [[Conceptual schema|conceptual model]] captures high-level business entities and their relationships, independent of any technical implementation; a [[Logical data model|logical model]] elaborates structure and relationships in technology-agnostic detail, specifying [[Data type|data types]], [[Primary key|keys]], and [[Cardinality (data modeling)|cardinalities]]; a [[Physical data model|physical model]] translates the logical design into implementation-specific structures for a particular [[Database management system|database management system]]. This three-tier framework is associated with the [[ANSI-SPARC architecture|ANSI-SPARC architecture]] and remains widely used in enterprise data practice. Several major modeling paradigms have emerged across the history of [[Database|database]] development. The [[Relational model|relational model]], introduced by [[Edgar F. Codd]] in 1970, organizes data into tables of rows and columns governed by [[Relational algebra|relational algebra]] and remains the dominant paradigm for [[Transactional database|transactional systems]]. [[Entity–relationship model|Entity–relationship modeling]], developed by [[Peter Chen]] in 1976, provides a graphical notation for representing entities, attributes, and relationships at the conceptual and logical levels. Alternative paradigms include [[Hierarchical database model|hierarchical]], [[Network model (database)|network]], [[Object-oriented database|object-oriented]], and [[Document-oriented database|document-oriented]] models, as well as purpose-built approaches such as [[Data vault modeling|Data Vault]] and [[Dimensional modeling|dimensional modeling]] for [[Data warehouse|data warehousing]]. The choice of data model has significant downstream consequences for [[Query language|query]] expressiveness, [[Data integrity|data integrity]], [[Scalability|scalability]], and [[Data governance|governance]].