Density: A measure of the diversity of concepts addressed in semantic networks

• We propose a novel methodology to grow up cliques networks.
• We propose density to measure the diversity of concepts addressed in semantic networks.
• We measure the Euclidean distance between different scientific journals.
• The density of clique networks may be of interest for the definition of new bibliometric parameters.

In this paper, we studied density effects in semantic networks constructed from a database of titles of papers published in scientific journals as a parameter to indicate the diversity of concepts in a journal. The proposed method essentially consists of fixing the number of titles for all of the studied scientific journals and analyzing the behavior of the density variation curves with regard to the inclusion of cliques (that is, complete networks associated with the titles). We observed that density behaves as a critically self-organized object when titles (cliques) are included in the network.