Data Modeling (RDBMS or Code)

A data model in software engineering is an abstract model that try to represent the reality and describes how data are represented and accessed.

Data models formally define:

• data entity
• and relationships among them

Data modeling can be found:

• in RDMS where a table represents a data entity and the keys, the relationship
• in code where object (of classes) can represent a data entity and the relationship can be represented by an extra attributes or an extern map.

There is not a right or wrong way of doing things as:

• design decision involves trade-offs between:
  • performance
  • against reduction of data duplicate and design complexity.
• every data model is meant to evolve. See Gay marriage: the database engineering perspective

• Data Modeling - Entity (Object, Table)
• Relation - Column
• Data Modeling - Identifier - (Id|Primary|Natural|Surrogate|Business) key
• Data Modeling - Relationships
• Data Modeling - Database Normalization
• Data Modeling - Denormalization
  • Data Modeling - Denormalization Methods (Changing the Database Design to Reduce the Number of Joins)
• Data Modeling - Referential Integrity (cascade)
• Data Modeling - Cardinality
• Data Modeling - Degree of a relationship
• Data Modeling - Data dictionary
• Data Modeling - (Metadata|Descriptor)
• Data Modeling - Ontology

The difference between model (ER, Dimensional, Vault) has everything to do with terminology and method.

There are two basic types of physical models :

• Data Modeling - Entity-Relationship (E-R) (E-R) and

E-R models are designed to minimize data storage redundancy and optimize data updates.

• Dimensional Modeling - Dimensional Schemas

Dimensional models are designed to enhance understandability and to optimize query performance.

• Data Vault

The Data Vault consists of three primary entity types:

• Hubs,
• Links,
• and Satellites.

The Hubs are core business keys, the Links form all relationships, and the Satellites provide all descriptive information.

• Data_model