Structural Exploration of Chinese Character Combination Networks

Abstract

In this paper, we learn the statistical properties of Chinese character networks within the framework of complex network theory. We construct two Chinese character networks (CCN1 and CCN2) based on 7000 Modern Chinese characters in common use from different respects. In CCN1, a link between two nodes (each node correspond to one Chinese character) represents that one node is a component part of the other. In CCN2, a link represents that the two nodes can be combined into a new node. It is found that, both in CCN1 and CCN2, there are some common features such as small-world effect, scale-free behavior, hierarchical organization and disassortative mixing. These results effectively explain the origin and formation of Chinese characters, which is absolutely helpful in understanding and using Chinese characters.