Most of the existing studies focus on modelling an information diffusion in temporal dimension: from the traditional linear threshold model and independent cascade model to stochastic models. While just a few efforts have been made on understanding information diffusion in both temporal and spatial dimensions. Though these studies successfully characterize the spatiotemporal diffusion patterns by using partial differential equations (PDE) models or probabilistic models, they are limited in a specific type of diffusion. Due to the heterogeneity of human dynamics, the vast scale of complex interactions, and the diversities of diseases and information, spatiotemporal diffusion patterns demonstrate great variety. It is a challenge to capture the diverse spatiotemporal diffusion patterns with a generic model.
Commentaires