Modeling Propagation Momentum of Topics Based on their Participants' Brightness

For social networking services like micro-blog that require unidirectional weak authentication, the link quality and strength of social network demonstrate the characteristics of heterogeneity and distortion. Given the fact that user participation is one of the main impetuses to social network evolution, we present a measure called user brightness to evaluate user activeness which is subsequently used to calculate propagation momentum of specific topic. Based on data-driven approach, it attributes the change of user activeness to his active participation in multi-topics, and the social network evolution and the dynamic change of network topology to aggregated activeness of massive users. The algorithm is low overhead and supports incremental update. Experiments on real data retrieved from Chinese micro-blogs shows that the proposed method demonstrates the momentum of topics very well.