Topologies and Centralities of Replied Networks on Bulletin Board Systems

A new complex network model is studied on a Bulletin Board System (BBS) website. The model has as its vertices the identifications registered on the site, and an edge is linked if one article refers to another. Statistical characteristics of this network, such as degree distribution, average shortest-path length, and clustering coefficient are investigated. Centralities of this network including degree, betweenness, and closeness are presented. The model displays small-world characteristics with small average shortest-path lengths and a high clustering coefficient. It also exhibits scale-free topology but the exponent of power-law distribution does not exceed 2. The activity, control, and independence of communication on this network are stronger than that of some social networks. The phenomenon can be explained by the fact that the communication of viewpoints is very fast, and one person's change of view has a significant influence on other people using the BBS.