Abstract:In this paper Chinese phrases are modeled using complex networks theory. We analyze statistical properties of the networks and find that phrase networks display some important features: not only small world and the power-law distribution, but also hierarchical structure and disassortative mixing. These statistical traits display the global organization of Chinese phrases. The origin and formation of such traits are analyzed from a macroscopic Chinese culture and philosophy perspective. It is interesting to find that Chinese culture and philosophy may shape the formation and structure of Chinese phrases.To uncover the structural design principles of networks, network motif patterns are studied. It is shown that they serve as basic building blocks to form the whole phrase networks, especially triad 38 (feed forward loop) plays a more important role in forming most of the phrases and other motifs. The distinct structure may not only keep the networks stable and robust, but also be helpful for information processing.The results of the paper can give some insight into Chinese language learning and language acquisition. It strengthens the idea that learning the phrases helps to understand Chinese culture. On the other side, understanding Chinese culture and philosophy does help to learn Chinese phrases. The hub nodes in the networks show the close relationship with Chinese culture and philosophy. Learning or teaching the hub characters, hub-linking phrases and phrases which are meaning related based on motif feature should be very useful and important for Chinese learning and acquisition.

Chinese Lexical Networks: the Structure, Function and Formation