Decision Support Tools for Ontological Engineering - Efficient Decision Support Systems: Practice and Challenges from Current to Future - page 57

b. If the ontology includes both static and dynamic knowledge, the user provides the

paths to both the static and dynamic knowledge files to Onto3DViz so that it can
import the ontology from both paths.

2. Extraction and Interpolation of Knowledge

a. Onto3DViz extracts and interpolates the OWL file(s) by applying a Protégé-OWL

API. The Protégé-OWL API originally is one of the plug-in packages from the
Protégé-OWL editor. This API is efficient in extracting and interpolating
knowledge from an OWL file.

b. Onto3DViz parses an XML-formatted file that stores dynamic knowledge by using

Document Object Model (DOM 4 ).

3. After the knowledge has been extracted from the ontology, it is stored into the Java

objects that are custom designed for storing visual data in the computer memory, and
which are used by Onto3DViz for creating the 3D model.

3.1.2 Visual object creation and display
The objective of Onto3DViz is to visually represent knowledge elements specified in the
IMT. The tool employs three techniques for accomplishing this objective; they are (1)
representation of concepts by the shape of visual objects, (2) size scaling of visual objects,
and (3) positioning of visual objects. These techniques enable Onto3DViz to represent an
ontology with minimal literal description while enhancing visual effects of the
representation. The three techniques are discussed in detail as follows.

3.1.3 Shape of visual objects
Onto3DViz uses the following symbols to represent the following concepts:

 Class : Sphere

 Objective : Cylinder

 Instance : Box

 Task : Cone

 Relationship between concepts : Line

3.1.4 Size scaling of visual objects
For the purpose of showing hierarchical structure among concepts of an ontology,
Onto3DViz decreases the size of the visual objects that occupy the lower levels in the
hierarchy. For example, if the class of Thing is the parent class of all classes, then it is scaled
to be the largest sized object. A subclass of the Thing class is a class, and the visual object of
this class is scaled to be 70 percent of the size of the Thing class. Similarly, subsequent
children classes are scaled to be 70 percent of the associated parent class. Hence the classes
in the default hierarchical view of an application ontology represents the visual objects that
are decreasing in size from top to bottom. By employing this technique, the hierarchical
relationships as reflected in successive parent-and-child relationships are clearly
represented among concepts.
4 Document Object Model, http://www.w3.org/DOM/