Original Date: 11/03/1996
Revision Date: 01/18/2007

Oak Ridge National Laboratory’s (ORNL’s) integrated training (Figure 3-1) exists in a variety of learning environments and applications. These applications range from simple browsers to adaptive and intelligent tutoring systems. ORNL is currently working on intelligent tutoring systems which can separate the knowledge base from the training application. By separating the knowledge base from the application, learning aids can be developed incrementally. Over time, the initial knowledge base can be expanded without affecting the training applications. This reusable training data approach allows expert knowledge to be captured as it is identified, without specifying exact requirements or delaying the development of an end product because of resource shortages. In addition, this approach works regardless of whether the eventual system becomes a browser, a tutorial, an expert, or a combination of applications.

Through the reusable training data approach, ORNL first gathers the expert knowledge for the database and then builds access methods to meet the users’ needs. Figure 3-2 illustrates ORNL’s approach to reusable training data. Today's databases can handle text, images, audio clips, graphics, and limited video. Modifications to the data need not affect the application. Applications can be written in hypertext markup language, authoring systems, programming language, or in the database’s own application. Numerous applications are possible with a single database and can serve many purposes including reference, learning, help, certification, and instructional feedback. In addition, applications can be housed on a CD-ROM or a server, and can be accessed from a stand-alone PC, a local network, or over the Internet. The ease of configuration management and updating the knowledge base makes Internet and networked applications particularly attractive.

For more information see the Point of Contact for this survey.