Data Models in DBMS

 

Data Models 

·        • Data models are frameworks or blueprints that describe how data is organized, stored, accessed, and managed in databases. 

• They provide a structure for representing data, relationships, and constraints, ensuring efficient management, reliability, and integrity of data. 

 

Common features

 Abstraction of complex real-world scenarios,

·   Enforcement of data integrity and relationships,

·   Support for data operations such as insertion, deletion, updating, and querying. 

 

Levels of Data Models

 

• Conceptual Data Model: 

  - High-level abstraction, not tied to implementation. 

  - Shows entities, their attributes, and general relationships. 

  - Visualized with ER diagrams. 

 

• Logical Data Model: 

  - Details the data structure (tables, columns, normalization, constraints) without physical storage specifics. 

  - Used to create blueprints for database design and ensure data integrity.

 

• Physical Data Model: 

  - Concerns about how data is actually stored on disk, including indexes, file placement, partitions, and storage parameters. 

  - Used for implementation and performance tuning.

 

Characteristics and Benefits

 • Abstraction hides implementation complexity, allowing for easier management and understanding. 

• Well-defined structures help maintain data integrity, consistency, and organization. 

• Operations are supported for managing and querying data. 

• Constraints, like primary keys and data types, enforce rules and relationships. 

• Data models help standardize communication between developers, designers, and stakeholders. 

• Scalability and flexibility enable databases to handle future changes and growth. 

Main Data Model Types

 1.    Hierarchical Data Models

• Organizes data in a tree-like structure (single root, parent-child relationships, each child has only one parent). 

• Best for data with a clear hierarchy (e.g., org charts, file systems). 

• Advantages: Simple, straightforward for 1:N relationships, efficient for certain queries. 

• Drawbacks: Rigid, inefficient for many-to-many relations or frequent changes, and redundancy issues.  

 

=================================================

 

2. Network Data Models

·         Network model was standardized as the CODASYL & DBTG (conference of data system language, database task group).

·        

Represents data as a graph: records (nodes) and sets (edges), allowing multiple relationships (M:N, many-to-many). 

• Suited for complex, interconnected data (telecom, transport). 

• Advantages: Flexible, efficient for traversals, handles data integrity with owner-member relationships. 

• Drawbacks: Complex to design and maintain, requires expert knowledge for querying and changes.

=================================================

3. Relational Data Models

 

 • Organizes data in tables (relations), where each row is a record and each column is an attribute. 

• Relationships handled using keys (primary, foreign). 

• Supports powerful querying (SQL), normalization to reduce redundancy. 

• Advantages: Easy to understand and update, ensures consistency, flexible with schema changes, widely used in business and transactional systems. 

• Drawbacks: May be less efficient at a massive scale or for unstructured data, joins can impact performance, and schema updates require planning.

=================================================

Entity-Relationship (ER) Model

• Used at the conceptual design phase.

Visualizes data as entities (objects), attributes, and relationships, often through ER diagrams. 

• Clarifies the high-level organization before implementation. 

• Not directly implemented, but serves as a foundation for creating relational databases.

Object-Oriented Data Model

• Integrates object-oriented programming principles (encapsulation, inheritance) with database storage. 

• Entities are objects, containing both data (attributes) and methods (behaviours).

• Useful for multimedia or engineering data, or where OOP and database need to work together. 

• Advantages: Manages complex data and structures, reusability, and encapsulation. 

• Drawbacks: Complex, not well-supported in some DBMS, querying can be harder.

NoSQL Data Models 

Designed for handling large-scale, flexible, and often unstructured data, usually in distributed systems.

Main types include: 

  • Key-Value: Simple pairs (e.g., Redis) – best for fast, scalable, simple lookups. 

  • Document: Schema-less documents (JSON/BSON, e.g., MongoDB) – flexible, supports complex structures. 

  • Column-Family: Organizes data into dynamic columns (e.g., Cassandra) – great for analytics and time-series data. 

  • Graph: Nodes and relationships (e.g., Neo4j) – best for highly interconnected data, such as social networks.

• Advantages: Scalability, flexibility, rapid lookups, schema-less storage. 

• Drawbacks: Weak consistency (eventually over strong), limited standards, querying complexity varies by model.

Usage Scenarios

• Hierarchical: File systems, org charts, manufacturing bill of materials. 

• Network: Telecom, transport, manufacturing, complex GIS data. 

• Relational: Business, finance, healthcare, customer management, e-commerce. 

• Object-Oriented: Multimedia, scientific, and spaces requiring OOP–DB integration. 

• NoSQL: Caching, real-time analytics, IoT, e-commerce, social media, big data.

==================================================================