Assessing the relevance of individual characteristics for the structure of similarity networks in new social strata in Shanghai

Node characteristics are found to be a key ingredient to shape the link formation in complex networks. In this paper, we perform a detailed investigation on the influence of different node characteristics on the structure of similarity networks, which are constructed based on a unique data set of questionnaires. The basic personal information (including the gender, experience of studying abroad, religion, and registered permanent residence in Shanghai (Shanghai Hukou)) and social positions (including intention of solving social problems, economic wealth, social reputation, and political status) provided in the questionnaires are considered as node characteristics. The similarity networks are built according to the Euclidean distances calculated based on the answer vectors to the questions on personal traits. According to the indicator Θ based on entropy measures, in 87% of tests, the links within the nodes having the same characteristics and between the nodes possessing different characteristics are not randomly connected. We further find that the node characteristics have significant effects on the preference of node in forming links within similarity networks. Our work sheds a new light on understanding the relevance of node characteristics for network structures.