Semantic networks were developed by those working in the area of artificial intelligence. The basic structures of this model consists of nodes and arcs, forming a network (a graph). The aim of these networks is the organization and representation of general knowledge about the world.
The initial target for the development of semantic networks was the natural language understanding, rather than data classification. Another feature of semantic networks is that there are as many as the needs that have different researchers in different projects.
Thus, it is difficult to decide which is called Bespoke Software Semantic Web data model. This is because they have been taken different data models that are good semantic network to represent a specific reality. We can say then that any graph in which nodes are connected by arches may be called a semantic network, where nodes and arcs are labeled.
To be a semantic graph, we need to define carefully the meaning of nodes and arcs, and how they are used [Moreno, 2000].
The first semantic networks used different nodes and arcs to represent the associations present in human memory. These early networks were not uniform in structure, failing to distinguish adequately between different types of nodes and arcs. For example, individual objects (instances) and object classes (entities) coexisted in the same semantic network. There was no clear difference between the nodes denoting denoting instances and classes, for example, the following semantic networks.
Basic Semantic Networks
Features.
a. Differentiate between types of object instances. Thus, classification is called the process of going from instances of objects to object types.
b. It introduces the concept of semantic distance, number of arcs that separate a node from another. In other data models this distance only has implications on performance, and generally not considered, nor has any semantic connotation [Jiménez, 2002].
In the distance semantic networks may be important, and is used to locate objects or closely related recently, depending on distance. In some cases the distance can be decreased by adding arcs for this purpose.
Semantic Distance
c. In semantic networks also have the idea of partition: the context of a network, in the sense of having a subnet, and for a specific task or work only part of the network is available. This facility is useful at the time of searching, and limiting the search space.
d. It also has the hierarchy of the type (or object). The types of hierarchies that have a semantic network are PART-OF and IS-A. The existence of a hierarchy implies that allows inheritance where an object is in a class that inherits all the properties of the class.
Heredity does not refer to inheritance of attributes and values, but also inherits the types of relationships allowed for that class. Example: If an employee is a person, and the relation “married” is valid for one, then also valid for engineers, lawyers and secretaries.