Please use this identifier to cite or link to this item:
|Title:||The Impact of Structural Characteristics of Concept Maps on Automatic Quality Measurement|
|Authors:||Hoppe, H. Ulrich|
|Publisher:||International Society of the Learning Sciences (ISLS)|
|Citation:||Hoppe, H. U., Engler, J., & Weinbrenner, S. (2012). The Impact of Structural Characteristics of Concept Maps on Automatic Quality Measurement. In van Aalst, J., Thompson, K., Jacobson, M. J., & Reimann, P. (Eds.), The Future of Learning: Proceedings of the 10th International Conference of the Learning Sciences (ICLS 2012) – Volume 1, Full Papers (pp. 291-298). Sydney, NSW, AUSTRALIA: International Society of the Learning Sciences.|
|Abstract:||Concept maps are often used in early phases of learning scenarios to externalise and possibly develop the conceptual understanding of a certain domain. An assessment of such maps can help to detect misconceptions and can serve as a basis for learner feedback. If used as a basis for agent-based scaffolding, the assessment needs to be calculated automatically by the system at run-time. Previous approaches have used an expert concept map as a reference for calculating such quality measures. This paper describes an approach that works without an explicit expert concept map but only uses more general domain ontology. To determine and adjust an automated quality measure student concept maps were collected and analysed by experts. The expert judgment was compared to certain structural measures based on graph-theoretical concepts and to enriched measures that additionally included ontology relations. It turned out that the inherent characteristic of concept maps as scale-free networks rules out certain structural measures as positive quality indicators.|
|Appears in Collections:||ICLS 2012|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.