Automated Scaffolding of Task - Based Learning in Non - Linear Game Environments

THOMAS, JAMES M. Automated Scaffolding of Task-Based Learning in Non-Linear Game Environments. (Under the direction of R. Michael Young.) This work described in this dissertation is an attempt to integrate intelligent tutoring capabilities into a framework that can be applied within any exploratory learning game. The goal is to provide dynamic adaptation to the learning needs of an individual students without constraining the autonomy and fun that digital games can offer. To accomplish this goal, a theoretical framework was designed that employs a novel planbased knowledge representation to describe both how the game works and what a student can learn in the game environment. This framework was implemented in a system called Annie, that employs a decompositional partial-order planner as its engine. Automated planning has been shown to provide a balance between user autonomy and story coherence within interactive narrative that is similar to the balance between player autonomy and learning progression that is a goal for Annie. Annie was implemented and evaluated with human learners in a game called FixIt, also created as a component of work for this dissertation. This document describes the theoretical framework, the implementations of Annie and FixIt, the methods and results of the experimental evaluations. c © Copyright 2011 by James M. Thomas