RuleWorks

A Framework for Model-Based Adaptive Training

Scheduling System - Procedural Knowledge

Routine procedures arise when enough knowledge is available to determine (without search) the correct step to take at any point in performing a task. The most efficient process to accomplish a task is to apply a given or standard procedure. This requires identifying the most appropriate procedure, but from then on the solution follows a specific path. No searching for alternative solutions is made; merely a pre-set sequence of operations is followed.

This is the most efficient but is also highly specific to a given situation. If the observed situation, or question during the training session, does not correspond to a standard procedure the trainee, or expert, must resort to the more general rules or indeed to the first principles [ITSIE 1992, D7]. In Scheduling-MOBAT fixed procedures can be attached to all tasks.

The technique used in Scheduling-MOBAT is a sequence of steps attached to appropriate procedural tasks. A simple two step procedural representation for Scheduling-MOBAT is shown in Figure 5-7 below. The “clash-process” procedure implements the subject specific task (to detect (clash)) which is mapped to the generic task (to identify (clash)) with a fixed procedure

Figure 5-7 Procedural Representation

Figure 5-7 Procedural Representation

At each of the procedural steps, further aspects to control a process operation are represented, such as timing aspects, conditional branching steps and specific actions to be taken. The representation of a fixed procedure is simply the embodiment of a set of task-dependent steps.

When a scheduling problem-space is entered, the system is located at an initial state for solving a part of the scheduling problem. The goal for entering a problem space is achieved when a desired state is reached. Selecting and applying a task with an appropriate fixed procedure is therefore one way of converting an initial state into a desired state.

This initial stage of starting a training task and proposing a way to solve a problem can be seen as identifying the most appropriate model of the domain for a given training objective and preferred trainee learning mode.

*
Flower Show
  *  
© RuleWorks.co.uk | | Sitemap