1 Sources

Monday, 6 May, 2024 4:11 AM

Data Models

Definition
Data Technique for defining business
modelling requirements for a database
Data • Simple representation of
models complex real-world data
structures
• Define how data is connected Reasons identifying business rules
to e/o and stored in the system • Help standardize company's view of data
Model Abstraction of real world object • Communication tool between designers and
or event users
• Allow designer to:
○ Understand nature, role, scope of data,
Importances
business process
Communication tool ○ Develop appropriate relationship
Overall view of the database ○ Create accurate data model
Organize data for various users
Translate
Abstraction for the creation of good • Nouns > Entities
database • Verbs > Relationship

Basic building blocks Naming conventions

Entity names Required to:
• Be descriptive
• Use terminology
Attribute Required to:
name • Be descriptive of the data
represented
Proper • Facilities communication between
naming parties
• Promotes self-documentation

Evolution of Data Models

Business Rules

Definition
Brief, precise, unambiguous descriptions of a
policy, procedure or principle
Enable defining the basic building blocks
Describe main and distinguish data
characteristics

Chapter 2 Page 1
2 Standard database concepts from network model
Monday, 6 May, 2024 still used by modern data models

Hierarchical Model
Details
Manage large amounts of data for complex
manufacturing projects
Represented by an upside-down tree which
contains segments (file system's record type)
Depicts a set of (1:M) relationship

Advantages and disadvantages

Advantages Disadvantages Relational Model
Details
Promotes data sharing Requires knowledge
of physical data Produced an auto transmission database to
storage replace standard transmission database
characteristics Based on the relation
Parent/child Changes in structure Relation/ table Matrix composed of intersecting
relationship promotes require changes in all tuple & attribute
conceptual simplicity app programs Tuple Rows
and data integrity Attribute Columns
Database security Implementation Describes a precise set of data manipulation
provided by DBMS limitation constructs
Efficient (1:M) Lack of standards
Advantages and disadvantages
Network Models Advantages Disadvantages
Details Structural Requires substantial
Represent complex data relationships independence using hardware and system
independent table software overhead
Improve database performance and impose a
database standard Tabular view improves Conceptual simplicity
conceptual simplicity gives untrained people
Depicts both (1:M) and (M:N) relationships tools to use a good
system poorly
Advantages and disadvantages
Ad hoc query May promote info
Advantages Disadvantages capability is based on problems
Conceptual simplicity System complexity SQL
limits efficiency Isolate the end user
Handles more Navigational system from physical level
relationship types yields complex details
implementation
Data access is flexible RDBMS & SQL
RDBMS SQL
Perform basic End user interface: Allows
functions end user to interact with
data
Make relational data Collection of tables stored
model easier to in database, each table
understand independent
Hides the complexities SQL engine: Executes all
of the relational model queries
from user

Chapter 2 Page 2
3 Advantages and disadvantages
Monday, 6 May, 2024 Advantages Disadvantages
Sematic content Slow development
The Entity Relationship Model Visual representation Complex navigational
Details system
Graphical representation of entities &
Inheritance promotes Learning curve is
relationship in database structure
data integrity steep
ERD Uses graphic representation
to model database
Object/ relational & XML
components
Details
Entity instance/ Row in relational table Extended relational Supports OO features and
occurrence data model (ERDM) complex data
Connectivity Term for relationship types representation
label O/R DBMS Based on ERDM, focuses
on better data
Advantages and disadvantages management
Advantages Disadvantages Extensible Markup Manages unstructured
Visual modelling yields Limited constraint Language (XML) data for efficient and
conceptual simplicity representation effective exchange of all
data types
Visual representation Limited relationship
makes it an effective representation & no
communication tool data manipulation Big Data
language Details
• New and better ways to manage large
Integrated with the Loss of info content amounts of web and sensor-generated data
dominant relational occurs when • High performance & scalability at
model attributes are reasonable cost
removed from
entities to avoid Volume amount of data
crowded displays Velocity speed in data growth rapidly
Variety data comes in multiple different formats
Object-Oriented Data Model (OODM)/
Sematic Challenges & new technologies
Details
Challenge New technology
OOBBMS Based on OODM
Volume doesn’t allow Hadoop Distributed File
Object • Contains data/ relationships usage of conventional System (HDFS)
with operations structure
• Basic building block for
autonomous structures Expensive Hadoop
• Abstraction of real-world entity OLAP tool inconsistent MapReduce
Attribute Describe properties of object dealing with
unstructured data
Class Collection of similar object with
shared structure/ behavior NoSQL
• Class hierarchy: Resembles
upside-down tree, each
class has 1 parent only
Inheritance Object inherits methods/
attributes from parents
Unified Describes sets of
Modelling diagram/symbols to graphically
Language model a system
(UML)

Chapter 2 Page 3
4
Monday, 6 May, 2024 9:07 PM

NoSQL
Details
Not based on relational model
Support distributed database architecture
High scalability, availability and fault tolerance
Large amount of sparse data
Stored key-value stores
Performance > Transaction

Advantages and disadvantages

Advantages Disadvantages
High scalability, Complex programming
availability, fault
tolerance provided
Low cost commodity No relationship
hardware support
Support Big Data No transaction
integrity support
Key-value model Inconsistent data
improves storage consistency
efficiency

Data Abstraction Levels

Chapter 2 Page 4