Compared with network datasets, multi-dimensional data are much more common nowadays. If we can model multi-dimensional datasets into networks with accurate network properties, while, in the meantime, preserving the original dataset features, we can not only explore the dataset dynamic but also acquire abundant synthetic network data. This paper proposed a fast scale-free network model for large-scale multi-dimensional data not limited to the network domain. The proposed network model is dynamic and able to generate scale-free graphs within linear time regardless of the scale or field of the modeled dataset. We further argued that in a dynamic network where edge-generation probability represents influence, as the network evolves, that influence also decays. We demonstrated how this influence decay phenomenon is reflected in our model and provided a case study using the Global Terrorism Database.