Understanding the Role of Structure in Concept Maps

Concept mapping is widely used in educational settings to aid knowledge construction, sharing, and comparison, as well as to assess students’ understanding. However, despite this rich range of uses, there have been few controlled studies of the relationship between concept maps and people’s internal conceptualizations. Because concept maps represent information as a network of concepts, an interesting question is how structural factors relate to concept importance. This paper presents results on modeling map-builders’ assessments of concept importance in terms of structural factors. Twenty subjects were asked to construct a concept map on a topic of interest and subsequently to rank labels of concepts extracted from their maps, based on the concepts’ importance in describing the map’s topic. Analysis of the results supports that subjects rank concepts higher if they are closer to the map’s root concept and if they have more outgoing connections or incoming connections relative to the other concepts. Agreement between the model’s predictions and subjects’ rankings was high, with Spearman’s rank correlation coefficients in some cases 1.0 or close to 1.0. These results suggest that topology alone is a sufficient indicator to extract topic-relevant information from concept maps. This has ramifications for developing tools to assess student concept maps and concept map authoring support systems. The results have been applied in the design of ”suggester” systems that aid domain experts in building concept maps, by using structural information to identify important concepts and search for topic-relevant information in previously built concept map libraries and on the Web.