01-11-2012, 12:39 PM
Entity – Relationship Data Model
Entity – Relationship Data.pdf (Size: 536.48 KB / Downloads: 25)
Motives and Role
• P. P. Chen is considered to be the author of the Entity –
Relationship (ER) data model
• The ER model is introduced in 1976
• The basic motives to introduce a new data model were
to provide:
– A semantically rich data model,
– Graphical representation of concepts, and
– To combine good characteristics of the network, relational, and
entity sets models
• The model is enhanced by many other authors
• To day, EER data model is extensively used in the
process of database design
Attributes
• An attribute is a property of a set of similar UoD objects
• A domain of values is associated with each attribute
• Some attributes may get a null value, as well
• Attributes can be classified according to many criteria
• Simple (non decomposable) and composite
(decomposable) attributes
• e.g.
– FName (simple),
– Address (composite)
Relationship Type
• A relationship type is an abstract representation of an
association between two or more UoD entity sets, and a
representation of a set of corresponding relationships
• Graphically, a relationship is represented by a diamond
shaped box connecting associated entity types
Weak entity type
• There may exist an entity set whose entities are
identified by their association to entities of another class
• Entities of the first class are identification and,
consequently existentially dependent on the entities of
the second class
• The chain of identification dependency may be longer
then two
• The identification dependent entities are represented by
a weak entity type, and they are connected to their
"owners" by weak relationship types
• Both weak entity and relationship types are represented
by double lined shapes
Superclass / Subclass
Relationship
• A subclass contains only specific attributes and may
participate in relationship types by its own
• A subclass is a specialization of its superclass
• The superclass is generalization of all its subclasses
• Specialization and generalization are inverses of each
other
• An instance of the subclass is also an instance of the
superclass
• These two instances represent the same real entity
• A subclass instance inherits all the superclass attribute
values and all relationship participations from its image
in the superclass
Summary
• EER data model is introduced to provide more semantic
power to data modeling
• It introduces a number of new modeling constructs and a
diagrammatic technique
• Set of similar entities is represented by an entity type
• Set of associations between two or more entities is
represented by a relationship type
• Structural constraint has two components:
– Cardinality ratio, and
– Participation constraint
• Classification is represented by superclass / subclass
relationship
• Category is a subset of the union of two (or more)
superclasses