| Περίληψη |
Conceptual data models strive for a high degree of intuitiveness in order to be better used and understood by their human users. In naming, humans usually put concepts (the names of the concepts) into a specific context to convey a specific meaning. This context is the scope of every name. This thesis introduces and analyzes the name scope in semantic networks and especially in those described in the Telos language. The purpose of this thesis is to design a mechanism that provides an object with the ability to be in the scope of other objects. This means that the objects and their names are identified uniquely by other objects using a mechanism, which is called name scope mechanism. A semantic network described in Telos consists of two different kind of objects: objects that are nodes in the network, and objects that are links either between nodes or between links and nodes. Every object is distinguished by its identity, which can be: temporal, representational, local and/or global. In Telos, every object has a unique object identifier, which is assigned by the system. This is the global identity of the object and is used by the system. Additionally, every object has an atomic name given by the user. This name is the local identity of the object. Users determine objects using logical names, which are a combination of atomic names. Logical name is either the global or the local identity of the object depending on whether this name identifies this object uniquely or not. The nodes of a Telos semantic network have unique atomic names all over the network. Links are allowed to have the same atomic names if they originate from different objects. A name scope mechanism provides an object-node with the ability to be in the scope of another object. According to this, object-nodes can have the same atomic names if they are in the scope of different objects. In such cases, the logical name of the object is automatically produced by the combination of its atomic name and the name of the object, in the scope of which this object belongs. Three name scope mechanisms are defined for Telos: classification, generalization and attribution. According to the classification or generalization mechanism, an object is able to be in the scope of one and only one class or superclass of it, respectively. The attribution scope mechanism differs from the other two mechanisms because the links are objects. According to this mechanism, an object is able to be in the scope of one and only one other object via one and only one category (attribute class). Finally, an implementation of the name scope mechanism is proposed.
|
Εμβέλεια Ονόματος σε Σημασιολογικά Μοντέλα Δεδομένων
