A Database Management System (DBMS) is software that stores, manages and retrieves data from a database. DBMSs are classified into different types according to how they organise and store data ā that is, the data model they use. The four classic types are the hierarchical, network, relational and object-oriented DBMS. In modern computing, NoSQL databases have also become very important for handling large, unstructured data. This guide explains each type of DBMS with its structure, advantages and examples.
The four classic types of DBMS are hierarchical, network, relational and object-oriented.
Hierarchical DBMS stores data in a tree structure with one parent per child (e.g., IBM IMS).
Network DBMS uses a graph structure and supports many-to-many relationships (e.g., IDMS).
Relational DBMS (RDBMS) stores data in tables and uses SQL ā it is the most widely used type (e.g., MySQL, Oracle).
Object-oriented DBMS stores data as objects with attributes and methods (e.g., db4o).
The relational model was proposed by E.F. Codd.
NoSQL databases (document, key-value, column, graph) handle large unstructured data (e.g., MongoDB).
Structure: Data is organised in a tree-like (hierarchical) structure with parent and child records. Each child record has only one parent, but a parent can have many children ā a one-to-many relationship.
Features: ā¢ Data is accessed by moving down from the root, level by level. ā¢ Fast access for one-to-many relationships.
Disadvantages: ā¢ A child cannot have more than one parent, so many-to-many relationships are difficult. ā¢ Less flexible ā changing the structure is hard.
Example: IBM's Information Management System (IMS); the directory structure of an operating system; XML documents.
Structure: Data is organised as a graph, where a record can have many parent and many child records ā supporting many-to-many relationships. Records are connected by links called 'sets'.
Features: ā¢ More flexible than the hierarchical model. ā¢ Can represent complex many-to-many relationships.
Disadvantages: ā¢ The structure is complex and harder to design and maintain. ā¢ Navigating the data requires knowing the structure.
Example: Integrated Data Store (IDS); IDMS (Integrated Database Management System).
Structure: Data is stored in tables (called relations) made of rows (records/tuples) and columns (fields/attributes). Tables are linked using common fields (keys), such as primary keys and foreign keys. It is based on the relational model proposed by E.F. Codd.
Features: ā¢ Easy to understand and use; data is in simple tables. ā¢ Uses SQL (Structured Query Language) to manage data. ā¢ Supports data integrity, security and reduced redundancy (through normalisation).
This is the most widely used type of DBMS today.
Examples: MySQL, Oracle, Microsoft SQL Server, PostgreSQL, IBM Db2.
Object-Oriented DBMS (OODBMS): Structure: Data is stored in the form of objects, as used in object-oriented programming. Each object holds both data (attributes) and methods (functions). It supports features like classes, inheritance and encapsulation. Use: Good for complex data such as multimedia, CAD/CAM and engineering applications. Examples: db4o, ObjectDB, GemStone.
Object-Relational DBMS: A mix of relational and object-oriented features (e.g., PostgreSQL, Oracle with object features).
NoSQL DBMS: Structure: 'Not Only SQL' databases store large amounts of unstructured or semi-structured data without fixed tables. Main kinds are document, key-value, column-family and graph databases. Use: Big data, real-time web and cloud applications. Examples: MongoDB (document), Redis (key-value), Cassandra (column), Neo4j (graph).
The four classic types of DBMS are: hierarchical DBMS (tree structure), network DBMS (graph structure), relational DBMS (tables linked by keys), and object-oriented DBMS (data stored as objects). In addition, modern NoSQL databases (document, key-value, column-family and graph) are widely used for large, unstructured data.
In a hierarchical DBMS, data is arranged in a tree where each child record has only one parent (one-to-many relationships). In a network DBMS, data is arranged as a graph where a record can have many parents and many children, so it supports many-to-many relationships. The network model is more flexible but also more complex.
A relational DBMS (RDBMS) stores data in tables (relations) made of rows and columns, and links tables using keys. It is the most popular type because it is simple to understand, uses the powerful SQL language, reduces data redundancy through normalisation, and provides strong data integrity and security. Examples include MySQL, Oracle, SQL Server and PostgreSQL.
An object-oriented DBMS (OODBMS) stores data in the form of objects, just like in object-oriented programming. Each object contains both data (attributes) and methods (functions), and supports classes, inheritance and encapsulation. It is well suited to complex data such as multimedia and engineering (CAD/CAM) applications. Examples include db4o and ObjectDB.
What is Verbal Communication? (Types and Examples)
Learn the definition of Verbal Communication. Understand the difference between Oral and Written communication with advantages and daily examples.
Difference Between Compiler and Interpreter
Learn the key differences between a Compiler and an Interpreter. Understand how these language translators process code differently with examples.
Difference Between Hardware and Software ā Complete Comparison with Examples
Difference between hardware and software ā hardware is the physical part of a computer you can touch; software is the set of programs and instructions. Full comparison.
Difference Between LAN, MAN, and WAN Networks
Understand the differences between LAN (Local Area), MAN (Metropolitan), and WAN (Wide Area) computer networks. Compare their speed, range, and use cases.
Difference Between Linear Search and Binary Search
Linear search checks elements one by one (O(n)); binary search repeatedly halves a sorted list (O(log n)). Learn the full difference with examples and a comparison table.
Turn this guide into revision flashcards, a practice exam, or an AI-generated podcast ā free, no signup required.