Hierarchical activeness state evaluation model for BBS network community

Evaluating behavior state of online communities is of critical importance for the study of online forum systems. A quantitative hierarchical state evaluation model is proposed in this paper to evaluate activeness degree of network groups focusing on one common topic. It consists of three levels: behavior, individual, and group. And the computational method is developed based on the scale of a network group and the behavior of its individuals. The active state indexes of a group and its individuals are calculated by weighting the importance of individuals based on individual actions: posting or browsing. The experiment results show that this model can provide intuitive active status of network communities so that network opinion administrators are freed from tedious analysis tasks in order to have activeness states of different groups. A series of factors including time, title, and diffusion extent of a topic are illustrated to have effect on evolution of active state for network group behaviors. And we also illustrate that the Matthew effect exists in network communities. This model is valuable for guiding, managing and controlling network public opinions.