Database Reference
In-Depth Information
that is the main reason why the relational model (Codd, 1970) has been dominating among other data
models. And insufficient support of the set-oriented view on data is why object-orientated paradigm
is much less popular in data modeling than in programming. Concepts and inclusion allow us to turn
object data models into a set-based approach where any element is inherently a set and the notion of
set is supported at the very basic level. Also, the use of inclusion relation makes COM similar to the
hierarchical model (Tsichritzis et al, 1976).
Inclusion in the case of empty entity class can be used to extend values by adding additional fields
and describing a hierarchy of value types. For example, if we need to mark an amount with the date of
transaction then it can be described as follows:
CONCEPT
DatedAmount
IN
Amount
IDENTITY
Date date
ENTITY
// Empty
Of course, the same can be done using conventional classes but then we would need to have two
kinds of classes for values and for objects while concepts and inclusion describe
both
values and objects
types. In relational terms, this allows us to model
two
hierarchies of domains and relations using only
one construct (concept) and one relation (inclusion). In this sense, it is a step in the direction of unify-
ing object-oriented and relational models by uniting two orthogonal branches: domain modeling and
relational modeling.
PARTIAL ORDER
Logical Navigation
Concepts in COQL are partially ordered using the principle that dimension types are greater concepts.
For example, let us consider the following concept:
CONCEPT
Account
IDENTITY
CHAR(10) accNo
ENTITY
Person owner
Here concept Account has a dimension which represents its owner the type of which is Person. Ac-
cordingly, concept Person is a greater concept and concept Account is a lesser concept. The number of
greater concepts is equal to the number of dimensions (concept fields). And the number of lesser con-
cepts is equal to the number of uses of this concept as a dimension type. In diagrams, greater concepts
are positioned over lesser concepts.
Concept instances are also partially ordered using the following principle:
a referenced element is
greater than the referencing element
. Thus a reference always points to a greater element in the database.
For example, a bank account element (instance of concept Account) is less than the referenced account
