Property description and comparison of BLOG community at different scales

The characteristics of complex network are studied mostly at small-scale and large-scale, while there is still much to be done at mesoscale. Firstly, by using the comment data of a blog site, an undirected network is built. From the perspective of network topology, it is divided and defined into small-scale (micro-level), mesoscale (community structure) and large-scale (macro-level). According to the characteristics of users' communication and the factors which influence network emergence, the comment network is described from the aspects of connectivity, diversity, and activity. Furthermore, some quantitative metrics are selected to analyze the network characteristics at each scale, including network structure entropy, degree distribution, rich-club, and the association between clustering coefficient and degree. It is found that the statistics are similar between mesoscale and the other two scales, indicating that the large-scale and small-scale could be understood through the mesoscale.