A rule-based system for automatically evaluating student concept maps

Abstract

Concept maps have been heralded as an effective learning tool to help students integrate new concepts into their existing set of knowledge. They are also useful for evaluating student learning and helping to illuminate where learning has occurred and where invalid or incomplete ideas are held by the student. We have developed a web-based concept map construction and rule-based evaluation system called the Concept Mapping Tool (CMT) that is being deployed at the university level. After students use the drawing facility of CMT to construct individual concept maps for a particular topic that was presented in a course, they can then use the rule-based evaluation system to grade their concept maps against a criterion concept map created by the course instructor. Students are given immediate feedback on how to improve their concept maps, and they can use CMT iteratively to improve their understanding of the topic at hand. The rule-based evaluation or grading system is modeled in part on a manual system for the consistent scoring of concept maps. Our tests of the system show that there is a strong positive correlation (>0.80) between the scores on students’ concept maps given by the course instructor grading manually and by the CMT rule-based evaluation system.