Abstract
The article is concerned with the proposal of a new approach of conceptual database design, called evolving conceptual database design, which exploits the structure for the preservation of database classes/concepts within the design. In order to discuss the opportunity to take into consideration this approach, the structure is constructed starting from a database conceptual graph. The leaves of the structure are mapped to a logical/object database graph. Horizontal steps of constructive logical database design extend the model. The computational costs required to design the structure for the preservation of database classes/concepts, as well as the qualitative/conceptual costs of the logical models resulting from the constructive design, are discussed.
TopIntroduction
Research at boundary between information systems and software development is important for today’s technology; (Buragga & Zaman, 2013). Before creating an information system, the business and organizational domain in which the information system is used must be represented. This representation, called conceptual model, is a critical point for the success of the information system development.
Model-driven engineering is an approach to software development that separates the specification of the system functionalities from its implementation on a particular platform. Models are the primary artifacts of software engineering. Two main categories of mappings between models are distinguished:
- •
Vertical mapping, relating system models at different abstraction levels;
- •
Horizontal mapping, relating models at the same abstraction level.
The requirements for engineering information modeling enclose complex objects, data exchange, data sharing, web based applications, imprecision uncertainty, and knowledge management. Techniques, such as web and artificial Intelligence, have been introduced in industrial applications.
Databases and Database systems play a main role in supporting data management and in implementing engineering information. Conceptual database design is defined by approaches to mapping from conceptual to logical database models (Elmasri & Navathe, 2015). Algorithms of conceptual database graph partitioning (Locuratolo & Rabitti, 1998; Spagnolo, 2000), originally introduced as algorithms of mapping from semantic data models to object database systems, have been also exploited as techniques for constructive conceptual database design (Locuratolo, 2013). Information systems and software development is a dynamic, interesting research area, which provides the finest and fastest way to solve typical problems.
Innovative evaluation of quality is introduced to design a database conceptual model which can result from a vertical mapping or a horizontal mapping (Locuratolo, 2014a). In the former case, the cost of specifications/proofs is given through the cardinality of variables and constants; in the latter case, the conceptual model is evaluated in terms of hidden classes. These represent the saving that is obtained for what has not been explicitly specified/proven. The hidden classes can also be exploited to test other quality desiderata/dimensions, such as easiness of use and flexibility. The numeric saving of consistency can be determined from the corresponding conceptual evaluations. The vice-versa is not possible.
In this article, a new approach to conceptual database design, called evolving conceptual database design, is introduced. The approach integrates a structure for the preservation of classes/concepts within the conceptual database design: starting from a database conceptual graph, an ontology defined at the boundary between concept theory and computer science, called structure for the preservation of database classes/concepts, is designed (Locuratolo, 2015b); the leaves of this structure are mapped to classes of a logical database model. A constructive approach of logical database design is considered to handle some dynamic aspects of the conceptual database design and the quality of this approach is evaluated.
Key Terms in this Chapter
is-aC Relationship: Concept [ v ] is–a c concept [ u ] iff class ( u, U ) is-a class ( v, V ) in the conceptual graph G c .
Qualitative/Quantitative “Measure” of Quality: Class defined “subtracting” from the specialized logical class ( u , U e ) class ( u , U). Formally ( u , ( U e - U)) = ( u , U e ) - ( u , U). This class expresses the saving of implicit specification. From the qualitative “measure” of quality, the quantitative “measure” can be determined. The vice-versa is not possible.
Database Conceptual/Logical Class: Couple ( v, V ), where the former component is a label denoting the class objects, and connoting the corresponding concept and the latter component is defined by attributes associated with the objects.
is-a Relationship: Conceptual class ( u, U ) is in is-a relationship with conceptual class ( v, V ) iff: the objects of class ( u, U ) are enclosed into the objects of class ( v, V ); the attributes of class ( u, U ) are inherited from class ( v, V ); class ( u, U ) can have specific attributes.
Specialized Logical Class: Class ( u , U e ) determined from class ( u, U ) is-a o ( v, V ): each attributes of class ( v, V ) is inherited by class ( u , U e ) and each attribute of U e belongs to class (u, U) or to class ( v, V ).
Structure for the Preservation of Database Classes/Concepts: Ontology defined at the boundary between concept theory and computer science enclosing: all and only the classes/concepts related to a given conceptual graph G c ; all and only the is–a C relationships among classes/concepts ; all and only the incompatible concepts/ disjoint classes, which are leaves of the structure.
is-ao Relationship: Relationship differing from the is-a relationship , since the objects of class ( u, U ) and the objects of class ( v, V ) are disjoint.