BDM_Wk2 - DATA MODELS

Course Information

• Course Name: Business Data Management

• Semester: Spring 2025

• Lecture Timing: Tuesday and Thursday, 05:40 - 07:00 pm

• Location: BRR, Room: 2071, LIVINGSTON

• Instructor: Deepak Shinde

• Contact: ds1886@scarletmail.rutgers.edu

Project Phases

• Due Dates:

  • Team formation and topic selection (3-member team): 2nd week

  • Define entities, attributes, and relationships

  • Define keys and business rules covering integrity constraints

  • Create ERD (Entity Relationship Diagram)

  • Verify if the tables are in 3NF (Third Normal Form)

  • Generate data for tables

  • Develop 10 business scenarios for the project

  • Submit project report

  • Deliver project presentation

Important Links

• Normalization, Denormalization Video

• The Evolution of Data Models Video

• Data Abstraction Video

Data Models

Learning Objectives

• Understand data modeling and the importance of data models

• Learn the basic building blocks of data models

• Comprehend what business rules are and their influence on database design

• Explore the evolution of data models

• Identify emerging alternative data models and their needs

• Classify data models by their level of abstraction

Model and Data Model

Definitions

• Model:

  • An abstraction of a complex real-world object or event

  • Aids in understanding real-world complexities (e.g. faculty or student enrollment)

• Data Model:

  • A simple graphical representation of complex real-world data structures

  • Represents data structures, characteristics, relationships, constraints, transformations in DBMS

Data Model Examples

1. Data Structure

• Entities: Students, Programs, Courses, Enrollments, Instructors, Teaching Assignments

• Example Table:

  • Students table contains relevant student information

  • Enrollments table manages student-course relationships

2. Entities and Characteristics

• Unique identifiers:

  • Each student has a unique Student_ID

  • Enrollment records reflect student enrollments

3. Relationships

• Programs have multiple students

• Students enroll in multiple courses

• Instructors teach multiple courses, and courses belong to a program

4. Constraints

• Data Integrity: Maintained via constraints like NOT NULL, CHECK, UNIQUE

• Primary Keys (PK) and Foreign Keys (FK) enforce referential integrity

Data Normalization and Aggregation

1. Normalization Stages

• First Normal Form (1NF): Atoms and no multivalued attributes

• Second Normal Form (2NF): Non-key attributes fully dependent on primary key

• Third Normal Form (3NF): Removes transitive dependencies

2. Aggregation

• Utilize aggregate functions (SUM, AVG, COUNT) to generate reports on data

Data Modeling Process

• An iterative process creating specific data models based on problem domains

• Combination of textual descriptions and diagrams

• Must include data structure descriptions, rules for integrity, and data manipulation methodologies

Importance of Data Models

1. Structured Representation:

   • Helps understand data and interconnections

2. Data Integrity:

   • Enforces consistency and reduces anomalies

3. Blueprint for Developers:

   • Facilitates communication between stakeholders

4. Guidance for Structure:

   • Defines tables, columns, keys, and relationships

5. Good Practices for Scalability:

   • Supports efficient querying and governance

Business Rules

Overview

• Brief descriptions guiding actions within organizations

• Define entities, attributes, relationships, and constraints

• Must be documented and regularly updated

Sources of Business Rules

1. Company Management

2. Policy Makers

3. Department Managers

4. Written Documentation

Reasons for Documenting Business Rules

• Standardizes data views

• Aids designer understanding of data nature and role

• Provides clarity for changes ensuring accurate reporting and analytics

• Supports compliance with legal frameworks

Data Models Evolution

• Reflects changing data management requirements and technology

• Transition from hierarchical to complex flexible models

Relational Data Model

• Introduced by E. F. Codd, represents data in relations/tables

• RDBMS handles relational data simplifying user interaction

NoSQL Databases

Characteristics

• Non-relational model, supports distributed architecture

• High scalability, availability, and fault tolerance

• Suitable for sparse data collections

Advantages & Disadvantages

• Pros: Flexible schema, performance, adept at handling big data

• Cons: Requires complex programming, lacks traditional relationship support

Degrees of Data Abstraction

• External: End-user view, focuses on specific business operations

• Conceptual: High-level overview of data for the organization

• Internal: DBMS view, mapping to relational model constructs

Abstraction Illustrations

• Various figures depict external, conceptual, and internal relationships and frameworks for understanding data models.