The Statistic Properties of Chinese Semantic Network in HowNet

Abstract

We extract a network from HowNet, a semantic network based on relations of concept explanations, and calculate its statistic properties. We found it is a complex network with features of small-world and scale-free. Structure of semantic networks based on neighboring words in sentence, conceptual similarity and association also have features of small-world and scale-free. Most exponents of power law degree distribution of these networks have values near 3. This property leads to speculations that the evolving mechanism of these networks is preferential attachment. But our result shows the exponent of HowNet power law degree distribution is near 1, different from most of complex networks evolving by preferential attachment. This feature reminds us there are other factors besides preferential attachment needed to be taken into account. We speculate that aggregation is an important mechanism in the development of conceptual networks.