DBMS Notes

Data Models

A data model is a collection of conceptual tools for describing data, data relationships, data semantics, and consistency constraints. It provides an...

What is a Data Model?

A data model is a collection of conceptual tools for describing data, data relationships, data semantics, and consistency constraints. It provides an organized framework for thinking about and representing the structure of a database. Just as an architect uses blueprints to describe a building before construction begins, a database designer uses data models to describe the database before implementation.

A good data model answers three fundamental questions: What data do we store? How are different pieces of data related? What rules must the data satisfy?

1. Hierarchical Model

Data is organized in a tree structure with a single root and parent-child relationships. Each parent can have many children, but each child has exactly one parent.

│ ├── Student	Alice (CS)
│ ├── Student	Bob (CS)
│ └── Course	DBMS
│ ├── Student	David (Math)
│ └── Course	Linear Algebra
└── Student	Frank (Physics)

Data access: Navigation follows the tree — to find Alice, you traverse University → CS Dept → Alice. There is no way to jump directly; you always start from the root and navigate down.

Advantages:

Simple and fast for one-to-many (1:N) relationships
Efficient when access patterns follow the hierarchy (e.g., "all students in CS")
Low storage overhead

Disadvantages:

Cannot naturally represent many-to-many (M:N) relationships
Data redundancy: if Alice takes courses in two departments, her record is duplicated
Rigid structure: reorganizing the tree requires rewriting applications

Historical Example: IBM IMS (Information Management System), used since the 1960s for banking and airline systems.

2. Network Model

An extension of the hierarchical model that allows many-to-many relationships. Data is organized as a directed graph where records (nodes) connect through links (edges called "sets").

Student	Courses
Alice	CS101, MATH201, PHY301
Bob	CS101, MATH201
Carol	PHY301

Students can enroll in many courses, and each course can have many students.

Advantages:

Supports M:N relationships naturally
More flexible than hierarchical model
Reduced redundancy compared to hierarchical

Disadvantages:

Complex navigation: programmers must know the physical structure and follow pointers
Difficult to reorganize: changing the graph structure breaks applications
No declarative query language — all access is procedural

Historical Example: CODASYL (Conference on Data Systems Languages) databases like IDMS.

3. Relational Model

Proposed by E.F. Codd in 1970, the relational model represents data as a collection of tables (relations). Each table has rows (tuples) and columns (attributes). Relationships between tables are expressed through shared attribute values (foreign keys), not physical pointers.

sql example

-- Student table
+-----+-------+------+
| SID | Name  | Dept |
+-----+-------+------+
| 101 | Alice | CS   |
| 102 | Bob   | Math |
+-----+-------+------+

-- Enrollment table (M:N relationship via junction table)
+-----+--------+-------+
| SID | CrsID  | Grade |
+-----+--------+-------+
| 101 | CS101  | A     |
| 101 | MATH201| B+    |
| 102 | CS101  | A-    |
+-----+--------+-------+

Key Properties:

Data is accessed using a declarative language (SQL) — you specify *what* you want, not *how* to navigate
Physical storage is independent of logical structure
Mathematical foundation (set theory, predicate logic) enables formal optimization

Advantages:

Simple, intuitive table structure
Data independence (physical reorganization does not affect queries)
Powerful query language (SQL)
Strong theoretical foundation
Supports integrity constraints declaratively

Disadvantages:

Performance overhead for complex joins on very large datasets
Not ideal for hierarchical or graph-shaped data
Object-relational impedance mismatch with OOP languages

Examples: Oracle, MySQL, PostgreSQL, SQL Server, SQLite.

4. Entity-Relationship (ER) Model

The ER model is a conceptual data model used during the design phase. It represents real-world scenarios as entities (things), attributes (properties), and relationships (associations).

The ER model is not used for implementation directly — it is converted (mapped) into a relational schema.

5. Object-Oriented Model

Combines database capabilities with object-oriented programming concepts:

Objects have state (attributes) and behavior (methods)
Classes define object templates
Inheritance allows subclasses to extend superclasses
Object identity — each object has a unique OID independent of its attribute values

Class	Person
attributes	name, address
methods	getName(), setAddress()
Class	Student extends Person
attributes	GPA, enrollmentYear
methods	calculateCGPA()

Use cases: CAD/CAM systems, multimedia databases, scientific data. Examples: ObjectDB, db4o.

6. Object-Relational Model

A hybrid that extends the relational model with object-oriented features: user-defined types, inheritance, methods, and nested tables — while keeping SQL as the query language.

sql example

-- User-defined type in PostgreSQL
CREATE TYPE address_type AS (
    street VARCHAR(100),
    city VARCHAR(50),
    zip CHAR(6)
);

CREATE TABLE Employee (
    emp_id INT PRIMARY KEY,
    name VARCHAR(100),
    home_address address_type  -- nested composite type
);

Examples: PostgreSQL (supports custom types, inheritance), Oracle (object types).

Comparison of Data Models

Feature	Hierarchical	Network	Relational	Object-Oriented
Structure	Tree	Graph	Tables	Objects/Classes
Relationships	1:N only	M:N	M:N via FK	References/OID
Query language	Procedural	Procedural	Declarative (SQL)	OQL
Data independence	Low	Low	High	Medium
Flexibility	Rigid	Moderate	High	High
Theoretical basis	None formal	CODASYL spec	Set theory	OOP theory
Current usage	Legacy systems	Rare	Dominant	Niche

Choosing the Right Model

In modern practice, the relational model dominates for transactional systems (OLTP). However, the choice depends on the application:

Relational: Business applications, ERP, banking, e-commerce
Document/NoSQL: Content management, real-time analytics, flexible schemas
Graph: Social networks, recommendation engines, fraud detection
Object-oriented: Engineering/CAD, genomics, multimedia

Understanding all models gives you the vocabulary to evaluate database systems and pick the right tool for each problem.

Exam Focus

Revise definitions, diagrams, examples, and short-answer points for Data Models.

Interview Use

Prepare one clear explanation, one practical example, and one common mistake for this Database Management Systems (DBMS) topic.

Search Terms

dbms, database management systems (dbms), unit, data, models, data models

Related Database Management Systems (DBMS) Topics

Continue learning this concept

Unit 1Data IndependenceData independence is the ability to modify the definition or organization of data at one level of the database system without requiring changes to the...Unit 1Unit 1 — Introduction to DBMS & Entity-Relationship ModelThis unit lays the foundation for your entire understanding of database management systems. It introduces two major areas: the fundamental concepts of DBMS what databases Unit 1Advantages and Disadvantages of DBMSIn a traditional file-processing system, each department maintains its own files. The HR department might store employee names and addresses, and so does...Unit 1Aggregation — Database Management SystemsAggregation is an abstraction mechanism in the Extended ER (EER) model that allows a relationship set to be treated as a higher-level entity set. This makes...Unit 1Characteristics of DBMSA DBMS has distinct characteristics that differentiate it from simple file-storage systems. These characteristics are what make a DBMS powerful, flexible,...Unit 1Database Users and AdministratorsA database system serves many different types of users, each with distinct skills, responsibilities, and levels of access. Understanding these roles is...

PreviousData Independence NextDatabase Users and Administrators

DBMS Notes

Data Models

A data model is a collection of conceptual tools for describing data, data relationships, data semantics, and consistency constraints. It provides an...

What is a Data Model?

A good data model answers three fundamental questions: What data do we store? How are different pieces of data related? What rules must the data satisfy?

1. Hierarchical Model

Data is organized in a tree structure with a single root and parent-child relationships. Each parent can have many children, but each child has exactly one parent.

│ ├── Student	Alice (CS)
│ ├── Student	Bob (CS)
│ └── Course	DBMS
│ ├── Student	David (Math)
│ └── Course	Linear Algebra
└── Student	Frank (Physics)

Data access: Navigation follows the tree — to find Alice, you traverse University → CS Dept → Alice. There is no way to jump directly; you always start from the root and navigate down.

Advantages:

Simple and fast for one-to-many (1:N) relationships
Efficient when access patterns follow the hierarchy (e.g., "all students in CS")
Low storage overhead

Disadvantages:

Cannot naturally represent many-to-many (M:N) relationships
Data redundancy: if Alice takes courses in two departments, her record is duplicated
Rigid structure: reorganizing the tree requires rewriting applications

Historical Example: IBM IMS (Information Management System), used since the 1960s for banking and airline systems.

2. Network Model

An extension of the hierarchical model that allows many-to-many relationships. Data is organized as a directed graph where records (nodes) connect through links (edges called "sets").

Student	Courses
Alice	CS101, MATH201, PHY301
Bob	CS101, MATH201
Carol	PHY301

Students can enroll in many courses, and each course can have many students.

Advantages:

Supports M:N relationships naturally
More flexible than hierarchical model
Reduced redundancy compared to hierarchical

Disadvantages:

Complex navigation: programmers must know the physical structure and follow pointers
Difficult to reorganize: changing the graph structure breaks applications
No declarative query language — all access is procedural

Historical Example: CODASYL (Conference on Data Systems Languages) databases like IDMS.

3. Relational Model

sql example

-- Student table
+-----+-------+------+
| SID | Name  | Dept |
+-----+-------+------+
| 101 | Alice | CS   |
| 102 | Bob   | Math |
+-----+-------+------+

-- Enrollment table (M:N relationship via junction table)
+-----+--------+-------+
| SID | CrsID  | Grade |
+-----+--------+-------+
| 101 | CS101  | A     |
| 101 | MATH201| B+    |
| 102 | CS101  | A-    |
+-----+--------+-------+

Key Properties:

Data is accessed using a declarative language (SQL) — you specify *what* you want, not *how* to navigate
Physical storage is independent of logical structure
Mathematical foundation (set theory, predicate logic) enables formal optimization

Advantages:

Simple, intuitive table structure
Data independence (physical reorganization does not affect queries)
Powerful query language (SQL)
Strong theoretical foundation
Supports integrity constraints declaratively

Disadvantages:

Performance overhead for complex joins on very large datasets
Not ideal for hierarchical or graph-shaped data
Object-relational impedance mismatch with OOP languages

Examples: Oracle, MySQL, PostgreSQL, SQL Server, SQLite.

4. Entity-Relationship (ER) Model

The ER model is a conceptual data model used during the design phase. It represents real-world scenarios as entities (things), attributes (properties), and relationships (associations).

The ER model is not used for implementation directly — it is converted (mapped) into a relational schema.

5. Object-Oriented Model

Combines database capabilities with object-oriented programming concepts:

Objects have state (attributes) and behavior (methods)
Classes define object templates
Inheritance allows subclasses to extend superclasses
Object identity — each object has a unique OID independent of its attribute values

Class	Person
attributes	name, address
methods	getName(), setAddress()
Class	Student extends Person
attributes	GPA, enrollmentYear
methods	calculateCGPA()

Use cases: CAD/CAM systems, multimedia databases, scientific data. Examples: ObjectDB, db4o.

6. Object-Relational Model

A hybrid that extends the relational model with object-oriented features: user-defined types, inheritance, methods, and nested tables — while keeping SQL as the query language.

sql example

-- User-defined type in PostgreSQL
CREATE TYPE address_type AS (
    street VARCHAR(100),
    city VARCHAR(50),
    zip CHAR(6)
);

CREATE TABLE Employee (
    emp_id INT PRIMARY KEY,
    name VARCHAR(100),
    home_address address_type  -- nested composite type
);

Examples: PostgreSQL (supports custom types, inheritance), Oracle (object types).

Comparison of Data Models

Feature	Hierarchical	Network	Relational	Object-Oriented
Structure	Tree	Graph	Tables	Objects/Classes
Relationships	1:N only	M:N	M:N via FK	References/OID
Query language	Procedural	Procedural	Declarative (SQL)	OQL
Data independence	Low	Low	High	Medium
Flexibility	Rigid	Moderate	High	High
Theoretical basis	None formal	CODASYL spec	Set theory	OOP theory
Current usage	Legacy systems	Rare	Dominant	Niche

Choosing the Right Model

In modern practice, the relational model dominates for transactional systems (OLTP). However, the choice depends on the application:

Relational: Business applications, ERP, banking, e-commerce
Document/NoSQL: Content management, real-time analytics, flexible schemas
Graph: Social networks, recommendation engines, fraud detection
Object-oriented: Engineering/CAD, genomics, multimedia

Understanding all models gives you the vocabulary to evaluate database systems and pick the right tool for each problem.

Exam Focus

Revise definitions, diagrams, examples, and short-answer points for Data Models.

Interview Use

Prepare one clear explanation, one practical example, and one common mistake for this Database Management Systems (DBMS) topic.

Search Terms

dbms, database management systems (dbms), unit, data, models, data models

Related Database Management Systems (DBMS) Topics

Continue learning this concept

PreviousData Independence NextDatabase Users and Administrators