Using semantic reference set of linking words for concept mapping in biology

Abstract

Inspired by the semantic network studies we propose additional conventions for choosing linking words and arrive at a Reference Set of semantically well-defined linking words drawn from the Knowledge Representation area of research in the domain of biology. Each linking word in the set is assigned a dimension: part-whole, class-inclusion, spatial-inclusion, function and attribution. We study expert representations by content analysis of biology texts at three levels of increasing subject complexity. We compare the linking words used in these representations with the Reference Set and find an increasing degree of proximity to the latter. This indicates that experts tend to use more well-defined linking words. Regarding this proximity as a characteristic of expertise, we can encourage novices to re-represent their concept maps using the linking words from the Reference Set. We discuss the implications of the approach for science education.