Evolving Training Scenarios with Measurable Variance in Learning Effects

One major cost driver in simulation-based training (SBT) and Intelligent Tutoring System (ITS) development is authoring scenario content. Effort is compounded when creating a catalog of scenarios to enable optimal scenario selection for the individual learner. Automated Scenario Generation (ASG) methods have been successful at creating variants of scenarios using procedural generation, evolution, or event templates with variable parameters. These methods typically describe boundaries within which content may vary, but do not describe how the variation will affect learning.