Data-driven method for assessing skill-opportunity recognition in open procedural problem solving environments

Abstract

Our research goal is to use data-driven methods to generate the basic functionalities of intelligent tutoring systems. In open procedural problem solving environments, the tutor gives users a goal with little to no restrictions on how to reach it. Knowledge components refer to not only skill application, but also applicable skill-opportunity recognition. Syntax and logic errors further confound the results with ambiguity in error detection. In this work, we present a domain independent method of assessing skill-opportunity recognition. The results of this method can be used to provide automatic feedback to users as well as to assess users problem solving abilities. © 2012 Springer-Verlag.