PART 1 • ARCHITECTURE & DATA MODELS

See the Layers Between a User and Stored Data

Learn how a DBMS separates user views, logical design and physical storage—and how different data models organize relationships.

Level 02 of 18Foundation75–100 minutesLevel 1 recommended

BY THE END, YOU CAN

Explain three levels of data abstraction.
Differentiate schema from instance.
Compare logical and physical independence.
Select a suitable data model.
Trace one-, two- and three-tier requests.

01 • UNDERSTAND

Abstraction Hides Detail Without Losing Meaning

A student checking a result should not need to know which disk page stores the marks.

STUDENT SEES

“My DBMS mark is 82.”

A simple view relevant to one user.

↔

DESIGNER SEES

Student — Enrolment — Course

The complete logical relationships and rules.

↔

SYSTEM SEES

Pages, records and indexes

The physical structures used for efficient access.

Schema

The relatively stable definition or blueprint: tables, fields, relationships and constraints.

STUDENT(student_id, name, branch)

Instance

The actual data stored at a particular moment. It changes whenever rows are inserted, updated or deleted.

(101, 'Anu', 'AIML')

02 • EXPLORE THE THREE-SCHEMA ARCHITECTURE

One Database, Three Levels of Description

Select each level to see who uses it, what it contains and what it deliberately hides.

External–Conceptual Mapping ↓

Conceptual–Internal Mapping ↓

Different external views of the same conceptual database

STUDENT PORTALname, enrolled_courses, own_results

FACULTY PORTALcourse, student_list, marks_entry

PLACEMENT CELLstudent, skills, eligibility, applications

03 • PREDICT CHANGE IMPACT

Data Independence Protects Higher Levels

Select a change and predict whether user programs should need modification.

PHYSICAL DATA INDEPENDENCE

Change storage without changing the logical schema

Examples: add an index, change file organization, move data to faster storage or alter page layout.

Usually easier to achieve

LOGICAL DATA INDEPENDENCE

Change the logical schema without breaking external views

Examples: split a table, add an attribute or introduce a new relationship while preserving user-facing views.

Usually harder to achieve

External views← protected by mappings →Conceptual schema← protected by mappings →Internal schema

SELECT A CHANGE

Which level changes, and which levels should remain stable?

Use the buttons to trace the effect.

04 • ORGANIZE THE WORLD

A Data Model Defines Structure, Relationships and Operations

A model is not merely a drawing style. It establishes how data is represented and connected.

CONCEPTUAL

High-level model

Describes the real-world meaning of data independently of implementation.

Example: ER model

LOGICAL / REPRESENTATIONAL

Implementation-ready structure

Describes records and relationships without exposing storage details.

Examples: relational, hierarchical, network

PHYSICAL

Storage-level model

Describes record layout, files, indexes, partitions and access paths.

Example: B+ tree index over disk pages

Model	Main structure	Relationship strength	Good fit	Main limitation
Hierarchical	Tree	Parent–child; normally one parent	Stable hierarchies and directory-like data	Many-to-many relationships are awkward
Network	Graph-like records and links	Multiple parents and many-to-many	Complex navigational relationships	Structural navigation makes programs tightly coupled
Relational	Tables	Keys and joins	Business systems and structured querying	Deeply nested or graph traversal can require many joins
Object-oriented	Objects and classes	References and inheritance	Complex objects closely matching application code	Less universal tooling and query conventions
Document	JSON-like documents	Embedding and references	Flexible, evolving aggregate records	Duplication and cross-document consistency need care
Key–value	Key mapped to opaque value	Mostly application-managed	Caches, sessions and very fast direct lookup	Limited ad-hoc querying and relationships

05 • APPLY

Choose a Model for the Requirement

There is no universally best model. Start from access patterns, relationships and consistency needs.

Application requirement

06 • TRACE REQUEST FLOW

One-Tier, Two-Tier and Three-Tier Architecture

“Tier” describes deployment separation. It is different from the three-schema levels of abstraction.

Important distinction: external/conceptual/internal are abstraction levels; one/two/three-tier are deployment arrangements. They answer different questions.

07 • CHECK YOUR UNDERSTANDING

Five Formative Concept Checks

Answers appear immediately because these questions teach rather than rank you.

1. Which level describes file organization and indexes?

2. Adding an index without changing tables demonstrates:

3. Which statement correctly distinguishes schema and instance?

4. Which model normally represents data as tables connected by keys?

5. In a common three-tier web system, where is business logic placed?

Answered correctly: 0 of 5

08 • EXPLAIN & PREPARE

University and Interview Questions

2-MARK QUESTIONS

Define schema and instance.
What is an external schema?
Define physical data independence.
Name three logical data models.
What is a two-tier architecture?

5-MARK QUESTIONS

Explain the three-schema architecture with a diagram.
Compare logical and physical data independence.
Compare hierarchical, network and relational models.
Differentiate one-, two- and three-tier architectures.

INTERVIEW QUESTIONS

Why is logical independence harder?
Give a real example of an external view.
Three-schema versus three-tier—what changes?
When would key–value storage be preferable?
What problem do mappings solve?

Show a strong answer framework: three-schema architecture

Define data abstraction and state its purpose.
Draw external views above one conceptual schema above one internal schema.
Explain what each level describes and hides.
Label both mappings between adjacent levels.
Connect the mappings to logical and physical data independence.
Use one continuous example such as a college database.

LEVEL 2 SUMMARY

You Can Now Read a Database Architecture

External views serve users; the conceptual schema describes the complete logical database; the internal schema describes storage.
Mappings translate between levels and support data independence.
Physical independence protects the conceptual schema; logical independence protects external views.
Data models organize structure, relationships, constraints and operations.
Three-schema abstraction and three-tier deployment are different concepts.

COURSE CHECKPOINT

Mark this level when you can draw the three-schema architecture and explain both forms of data independence without notes.

Saved in this browser only.