The process of designating sub-groupings within the entity set is called as specialization. Specialization allows us to distinguish among entities.

One limitation of E-R model is that it cannot express relationships among relationships. Thus for this purpose, aggregation is used. Aggregation is an abstraction through which relationships are treated as higher level entities.

Machine definition is not a generalization constraint. Condition defined, user defined, disjoint and overlapping are 4 generalization constraints.

We can apply specialization multiple times to refine a design. An entity set may also be specialized by more than one distinguishing feature.

The similarities between the entity set can be expressed by the generalization feature. It is a containment o the relationship that exists between a higher level entity set and one or more lower level entity sets.

Higher level entity sets can also be designated by the term super class. In the similar manner lower level entity sets can also be designated by the term sub class.

Condition defined generalization constraint is also said to be attribute defined.

If each higher level entity belongs to the lower level entity then it is called as total generalization. It is also called as total specialization. If some higher level entities do not belong, it is called as partial generalization.

Overlapping specialization is the type of specialization that permits multiple sets. But disjoint specialization does not permit multiple sets. Disjoint specialization permits at most one set.

The attributes of the higher level entity sets are inherited by the attributes of the lower level entity sets. But the inverse is not true in this case.