Approach to model WEB activity based on phase transition theory

Many kinds of interactive WEB spaces have attracted more and more people to express their own opinion, and communicate with others, etc. However, the complexity that is generated from topics, users¿ activity, and their evolution properties leads to the difficulty in describing the dynamic in WEB activities. Considering a WEB space as an open system which locates at a far from equilibrium state, we apply phase transition theory to capture the macroscopic dynamic in WEB activity and proposed a generative model. A Hurst sequence which identifies the macroscopic property like the temporal dependency and phase transition is defined. A set of Hurst sequences which is extracted from a WEB space is used to compute its dynamic characteristic. Related algorithms, such as FHS generating algorithm are provided. Experiments on FUDAN BBS analysis are done and the results show that the proposed method can identify the WEB dynamic well.