Graphing knowledge: automated concept map evaluation to assess and enhance student learning

Abstract

A concept map (CM) is a graphical representation of relationships and interconnections between concepts or ideas, using nodes and links to depict the hierarchical and interconnected structure of knowledge. A literature review on methods for evaluating students’ CMs was conducted, revealing potential areas for further research in CM evaluation techniques. To address these gaps, this study introduces a novel framework, graphing students’ knowledge using concept maps (GSKCM). GSKCM employs a graph-based scoring algorithm to automate the evaluation of concept maps and assess students’ structural comprehension, thereby facilitating targeted remediation. The algorithm compares student-generated CMs to a reference map, validating individual concepts and propositions while emphasizing the hierarchical structure— ultimately assessing students’ understanding of the topic’s overall framework. The GSKCM framework was implemented in an authentic classroom setting, where students constructed CMs on a specific topic. These maps were evaluated using the framework, revealing both strong and weak conceptual areas for individual students and the class as a whole. Based on the algorithm’s results, students were grouped into five clusters, enabling tailored tutoring at both the individual and group levels. An experimental study comparing the GSKCM approach to traditional evaluation methods demonstrated that students in the experimental group outperformed those in the control group on a formative assessment. By automating the evaluation process, GSKCM streamlines assessments, provides deeper insights into student understanding, and enhances the learning experience—particularly in mastering complex subjects.