Logical Data Modeling

Logical data modeling is the second phase of database design, translating a conceptual model into a detailed, DBMS-independent blueprint of the data structure. It defines entities, attributes, keys, relationships, and normal forms without specifying physical storage details.

What It Defines

Element	Description
Entities / Tables	Named objects with defined attributes
Attributes / Columns	Properties with [[SQL Data Types
[[Primary Key	Primary keys]]
[[Foreign Key	Foreign keys]]
Normalization	Decomposition to eliminate redundancy (typically to 3NF/BCNF)
Database Constraints	NOT NULL, UNIQUE, CHECK, referential actions

Expressed Through

• Entity-Relationship Diagram (ERD) — Crow’s Foot / IE notation is the standard for logical models
• SQL DDL — CREATE TABLE statements (though this overlaps with physical modeling)
• Data modeling tools — MySQL Workbench, pgModeler, LucidChart, ERDPlus

Conceptual → Logical Translation

The same conceptual model maps differently depending on the target database paradigm:

Conceptual Element	Relational Logical	Document Logical	Graph Logical
Entity	Table	Collection	Node
Relationship	Foreign Key / Junction Table	Embedded doc or reference	Edge
Attribute	Column	Field	Property
Generalization	Discriminator column or table-per-type	Type field	Label

Position in Design Workflow

graph LR
    A["[[Conceptual Data Modeling|Conceptual]]<br/>(UML / domain model)"] --> B["Logical<br/>(ERD / normalized schema)"]
    B --> C["[[Physical Data Model|Physical]]<br/>(DDL / indexes / partitions)"]