Catálogo de publicaciones - libros

In this chapter we review the basics of the conceptual modeling of information systems. We explain that conceptual modeling is a necessary activity in the development of an information system, the objective of which is to define the conceptual schema of the system. We also explain that conceptual modeling must be preceded and followed by other activities.

Determining the entity types that exist in a given domain and are relevant to an information system is a fundamental task in conceptual modeling. A clear understanding of entity types and their characteristics is therefore necessary. This is the focus of this chapter.

Relationship types are another important element in conceptual schemas, because they also play a fundamental role in the memory, informative, and active functions of information systems. Determining the relationship types that are relevant to an information system is one of the most important tasks in conceptual modeling. In this chapter, we study the nature and general characteristics of a relationship type.

Cardinality constraints are one of the most important kinds of constraint in conceptual modeling. In addition to constraining the population of relationship types, cardinality constraints help us to understand the meaning of the types involved, and they also play an important role in system design

As we saw in Chap. 2, “entity types are concepts whose instances are identifiable objects ...” An entity is identifiable if there is a linguistic expression that denotes it. Most of these expressions are built from reference relationship types, a particular kind of relationship type described in Sect. 5.1. In Sect. 5.2, we discuss how to use reference relationship types to identify entities. Section 5.3 explains that in general the participants of a relationship type should be entity types rather than its identifiers.

Reifying a relationship consists in viewing it as an entity. The word “reification” comes from the Latin word , which means “thing”. Reification has a well-known equivalent in natural language, nominalization, which basically consists in turning a verb into a noun. Reification is widely used in conceptual modeling; conceptual modelers must therefore have a good grasp of it. In Sect. 6.1, we define reification and explain its logical basis. Reification can easily be defined in UML, as we show in Sect. 6.2. In some languages, however, reification cannot be defined as easily, so one must instead use implicit reification, which is also described in Sect. 6.2. Implicit reification is an interesting schema transformation that can be used in other contexts.

In the four preceding chapters, we studied relationship types without taking into account their particular meaning. In general, the meanings of the relationship types existing in a schema are very diverse. However, there are some relationship types that appear in many schemas and even several times in the same schema. They are the subject of this chapter: in Sect. 7.1 we define them, and in Sect. 7.2 we show how they can be represented in an information system.

In this chapter we show that entity and relationship types may be base, derived, or hybrid (Sect. 8.1). The instances of base types need to be explicitly represented in an information base, while those of derived and hybrid types may be inferred by an information system, using derivation rules. Derivation rules are domain knowledge that an information system needs in order to derive certain facts; this knowledge must therefore be described in the conceptual schema. Section 8.2 describes the logical and the UML representations of derived and hybrid types and their derivation rules. In general, derivation rules are very diverse, although certain kinds appear very often. Section 8.3 describes some of these. Section 8.4 shows that the derivation rules of constant relationships require special interpretation. Section 8.5 explains how to define a particular kind of hybrid type in UML. Derived types add complexity to a schema, so their definition must be justified. Section 8.6 deals with the justification of derived types.

Section 9.1 studies the concept of an integrity constraint and its importance in conceptual modeling. Section 9.2 shows that integrity constraints can be classified from several points of view. These classifications help us in understanding the nature of integrity constraints. Section 9.3 describes the definition of static constraints in logic and in UML. In general, integrity constraints are very diverse, but there are some particular kinds that appear very often. Section 9.4 describes some of them. Section 9.5 identifies the creation-time constraint, an important particular kind of transition constraint, and explains a way to define it in conceptual schemas.

It is often the case that the instances of an entity type must also necessarily be an instance of another entity type. This can be understood as a special relationship, an relationship, between entity types (and, in general, between concepts). relationships are constraints. Entity types and their relationships form a network structure called a taxonomy. Taxonomies are a very important part of conceptual schemas, and the objective of this chapter is to study them.