Semantic Structuring of Concepts Found in Scholarly Literature: An Empirical Study Using a Fuzzy Clustering Approach

In this empirical study a concept map is reorganized and clustered in order to provide an intuitive structure for browsing. The concepts are represented as nodes created by the authors of articles but also by readers of these articles. Semantic dependencies are coded through links between the nodes to build a graph. We report an experiment in which such a concept graph is clustered through fuzzy clustering techniques in order to extract semantically dependent regions. These clusters can potentially act as new artificial concepts on a higher level of abstraction.