Knowledge Discovery from Growing Social Networks

摘要

The project explored mathematical models to explain, control and visualize a wide variety of information diffusion processes. The main results are the following six: 1) A very efficient method for minimizing the propagation of undesirable things by blocking a limited number of links in a network. 2) An effective visualization method for understanding a complex network, in particular its dynamical aspect such as information diffusion process. 3) A new scheme for empirical study to explore the behavioral characteristics of representative information diffusion models. 4) An effective method for ranking influential nodes in complex social networks by estimating diffusion probabilities from observed information diffusion data using the popular independent cascade (IC) model. 5) A very efficient method for discovering the influential nodes in a social network under the susceptible/infected/susceptible (SIS) model. 6) A new method for learning continuous-time information diffusion model for social behavioral data analysis.