DBMS Generalization, Specialization, and Aggregation

DBMS Generalization

Generalization is a bottom-up approach in which the common attributes of two or more lower-level entities combines to form a new higher-level entity. In generalization, the generalized entity of higher level can also combine with entities of the lower-level to make further higher-level entity.
It is like a superclass and subclass system, but the only difference is that it uses the bottom-up approach. It helps in reducing the schema size. It is always applied to the group of entities and result in the formation of a single entity.

According to the below diagram, there are two entities, namely teacher and student. The teacher entity contains attributes such as teacher_id, name, and address and student entity include student_id, name, and address. Both entities can be combined to create a higher-level entity person. The address and name are common to both the entities. The teacher entity has its attribute teacher_id, and the student has its attribute student_id. The entities teacher and student are generalized further into the person entity.
A lower-level entity is called a subclass, and the higher-level entity is called a superclass. So, the person entity is the superclass of two subclasses teacher and student.

DBMS Specialization

It is opposite or inverse of generalization. A specialization is a top-down approach in which an entity of higher-level entity is broken down into two or more entities of lower level. In specialization, a higher-level entity set may not have any lower-level entity set. It is always applied to a single entity and results in the formation of multiple new entities. It increases the size of schema due to the increase in the number of entities.

In the below example, it can be seen that the employee is a high-level entity which is divided into three sub-entities (Ram, Shyam, and Mohan). The sub-entities are the names of the employees (relating to the high-level entity). Therefore, splitting a high-level entity into a low-level entity is called specialization.

DBMS Aggregation

Through the Entity-Relationship model, we cannot express a relationship set with another relationship set. Thus we use the concept of aggregation to express it.

Aggregation is an abstraction in which the relationship between two entities is treated as a higher-level entity. It allows you to indicate that a relationship set participates in another relationship set.

Example:

The student and program are two entity set and related through a relationship set write. Suppose, a student who writes any program must require a programming language installed in their system. So, there will be another relationship set requires. You need to connect the relationship set requires with an entity set programming language and relationship set write. But, we can connect only entity set to a relationship set.
One relationship set cannot be connected with another relationship set; for this, we need aggregation. Here, the write (relationship set) will treat like a higher-level entity and can be associated with a relationship set requires. So, aggregation is needed when you express a relationship set with another relationship set.