Database Design

Database Models

Hierarchical

• Organizes data into tree structure.
• Parent and Child Data segments
• 1:N mapping between record types
• Child data node can have only one parent

Network

• Based on mathematical set theory
• More than one parent per child
• The "set construct": An owner record type, a set name, and a member record type.

Relational

• Defines data structures, storage and retrieval operations, integrity constraints
• Organized into tables with following properties:
  • Values are atomic
  • Each row is unique
  • Column values are of the same kind
  • Sequence of rows and columns is insignificant, i.e., interchangeable
  • each column has a unique name
• Designate certain fields as keys to optimize searching of that field
• join operation: take values from two different tables matched by common field
• Robust transaction and performance management features
• A complex data structure must be flattened out to fit into tables or joined together from those tables, which takes overhead

Object/Relational Model

• Adds new object storage capabilities to relational: traditional fielded data, plus complex objects
• SQL and Oracle are here

Object-Oriented Model

• Manip DB using C++, Java, no need to translate calls into SQL
• Less code, more natural data modelling.
• one-to-one mapping of DB objects and OO programming instances

Semi-structured Model

• Flexible, when you don't know the schema ahead of time.
• The schema is in the data, "self-describing"
• Modelled in terms of graphs which contain lables which give semantics to its underlying structure

Associative Model

• Divides data into entities (items), and associations (links)
• Source, verb and target