Generalization in database management system: 18 flashcards

Generalization is the process of defining a more general entity type from a set of lower-level entity types by identifying their common attributes. It is a bottom-up approach in conceptual design.

Generalization is a bottom-up process that combines multiple entity types into a higher-level supertype, while specialization is a top-down process that divides a higher-level entity type into lower-level subtypes based on distinguishing characteristics.

A triangle (or a circle with a line) connecting the superclass to the subclasses, often labeled with 'ISA' or a subset symbol.

A superclass is the higher-level, more general entity type that results from generalization. It contains attributes common to all its subclasses.

A subclass is a lower-level, more specific entity type that inherits attributes from the superclass and may have additional attributes or relationships of its own.

Attribute inheritance is a property where subclasses automatically acquire all attributes of their superclass. This avoids redundancy and ensures consistency.

The two main constraints are **disjointness** (whether a superclass instance can belong to only one subclass or multiple) and **completeness** (whether every superclass instance must belong to a subclass).

It means that an entity instance of the superclass can belong to at most one subclass. For example, a 'Person' can be either 'Student' or 'Employee', but not both.

It means that an entity instance of the superclass can belong to multiple subclasses simultaneously. For example, a 'Person' can be both 'Student' and 'Employee'.

It means that every superclass instance must be a member of at least one subclass. For example, every 'Vehicle' must be either 'Car' or 'Truck'.

It means that a superclass instance is not required to belong to any subclass. Some instances may exist only as the superclass type.

Combining 'UndergraduateStudent' and 'GraduateStudent' into a superclass 'Student' based on common attributes like StudentID, Name, and GPA.

By placing common attributes in the superclass, they are stored only once instead of being repeated in each subclass. This eliminates duplicate storage and reduces inconsistency risk.

The 'ISA' relationship (e.g., 'Student ISA Person') is used to represent generalization/specialization. It indicates that a subclass entity is a kind of the superclass entity.

Yes, generalization can be applied recursively, creating a hierarchy or lattice where a subclass can become a superclass for further specialized subclasses.

A hierarchy allows a subclass to have only one superclass (single inheritance), while a lattice allows a subclass to have multiple superclasses (multiple inheritance).

Create one table for the superclass that includes all attributes of the superclass and all subclasses, with a type discriminator column to indicate which subclass each row belongs to. This is simple but can lead to many null values.

Create a table for the superclass with its attributes, and separate tables for each subclass that include the subclass-specific attributes and a foreign key to the superclass table. This avoids nulls but requires joins.