Language as a Small World Network

Abstract

Small world networks are graphs, which integrate the properties of random graphs and lattice graphs [5]. Several real networks can be modeled with a help of small worlds [1, 2, 6, 8]. As shown in this paper a word web has small world structure too. We have shown, that different languages, such as Slovak and English, have similar small world properties. As our second goal, we built a graph on the basis of kernel lexicon words, in order to test the scaling results of Cancho and Sol [6]. We speculate that the differences between calculated and measured exponents of connectivity distribution are due to the node aging.