Implementing learner control in an automated instructional system

Abstract

In an automated instructional system, learner control suggests that there is some basis on which the system may recommend and the learner may select relevant and appropriate instructional elements and that the system has the capability to respond to learner requirements by customizing the delivery of each element. We report research into these two areas of sequencing and delivery. We first describe a knowledge representation model that expresses the interrelations among instructional elements. A domain knowledge base created using this model, available through all phases of the instructional development cycle, provides information to the system and the learner for identifying which elements are related in some way to the just presented element and, more importantly, the nature of the relation so that an informed sequencing decision may be made. We then describe transaction shells, reusable instructional components, which when instantiated with content from the domain knowledge base deliver instruction to the learner. This instruction is configurable both by the author and dynamically by the system to take into account knowledge about the learner, including aptitude, goals, and previous instruction. The automated system, by configuring a shell dynamically, adjusts an instructional element to meet expressed or derived learner requirements. These two elements combine to provide a foundation for implementing learner control in an automated instructional system.