Data Models Overview

Data Models

• A data model is a set of concepts used to:
  • Describe the structure of a database.
  • Define operations for manipulating data.
  • Enforce constraints on data within the database.
• Data Models contain constructs used to define a database's structure.
  • These constructs include individual elements (with data types) as well as groupings of elements (like entities, tables, or records)
  • Relationships between these groupings of elements are also defined.
• Constraints are limitations on the data that the database can store:
  • These must be observed throughout the database's lifetime.
• Data Models are used to define operations:
  • These operations allow for retrieving and updating data within the database.

Categories of Data Models

• Conceptual (semantic) data models are high-level and conceptual:
  • They are designed to be close to the way users perceive data.
• Physical data models are low-level and internal:
  • They describe physical data storage approaches.
• Implementation data models fall between the two.
  • They are used by many implemented commercial DBMS systems.
• Self-describing data models combine data descriptions and data values.
  • Examples include XML, key-value stores, and some NoSQL systems.

Schemas Versus Instances

• A database schema describes the structure of the database.
• It includes details such as data types and constraints.
• A schema diagram visually represents a database schema.
• A database state is the actual data in the database at any particular time.
• Database state is sometimes called the database instance and is a collection of all data.

Three-Schema Architecture

• The three-schema architecture was created to support:
  • Program-data independence: Changes in the physical schema should not cause applications to break.
  • Support for multiple views of the data: Different users have different needs and permissions.
• The architecture defines three levels of schemas:
  • Internal schema describes physical structures and access paths.
  • Conceptual schema describes the structure and constraints of the entire database.
  • External schemas define various user views.

DBMS Languages

• Data Definition Language (DDL) is used to define, alter, or delete database objects (Tables, Views, Indexes).
• Data Manipulation Language (DML) is used to retrieve and modify data in the database.
  • SQL is an example of a non-procedural high-level language often used with DML.
• Procedural languages are low-level and need to be embedded in a programming language.

History of Data Models

• The Network Model was first implemented in the 1960s.
  • It is based on records and set types and uses a navigational language.
• The Hierarchical model was popularized by IBM's IMS family of systems.
  • It represents data as a hierarchy and also utilizes a navigational language.
• The Relational Model was introduced by E.F.Codd in the 1970s and uses tables with rows (records) and columns (attributes).
  • SQL is the standard language for relational database systems.
• Object-oriented data models are based on data abstraction, objects, and inheritance concepts.
• Object-relational models combine aspects of relational and object-oriented models.
  • They have become a dominant way for database vendors to extend systems to support XML, text, and other data types.