Research On Group Spatial Patterns Of Specific Phenomena Under Networked Geographic Spaces

With the rapid development of the Internet, the network has become an indispensable tool for people to get information, express their views, or communicate with each other every day. The popularity of the Internet makes people involve in the discussion of specific events, therefore they spread very quickly. The dissemination of the network information flow of specific phenomena in a networked geographic space will result in unexpected spatial patterns. Understanding this issue will help to reveal the creating mode of the spatial patterns of group phenomena under networked geographic space.Based on the spatial correlation, the first law of geography, the spatial proximity, the spatial-temporal proximity and the cellular automata, considering the characteristics of network information flow, this paper gave the definition of spatial-temporal proximity of the networked geographic space, and proposed a mathematical equation for shaping spatial patterns of the specific phenomena under networked geographic spaces.Using the techniques of computer network, computer software and GIS, and the virtual network geospatial simulation model, based on the combination of views between the geographical cellular formations and the interactive view features, we built a networked geospatial cellular automata model. Based on Java Repast Simphony framework, we designed and implemented a networked geospatial cellular automata simulation platform; and carried out 16 experiments using the simulation platform. Then these experiments were performed statistical analysis and visualization analysis through Python Scikit-Learn machine learning module and Matplotlib mapping module. Finally, we established the mathematical equation for shaping spatial patterns of the specific phenomena under networked geographic spaces through experiment data.Conclusions of simulation experiments are described as follows:(1) There are five groups of cellular distribution in networked geographic spaces, which are the normal distribution, the single distribution, the bivariate distribution, the ternary distribution and the multivariate distribution. Each distribution is a reaction to the reality of social phenomena.(2) Geographical cellular community in networked geographic spaces shows aggregation phenomenon in spatial-temporal patterns. The evolution result of the geospatial cellular automata machine indicates that geographical cellular aggregation groups and real-life groups are highly similar.(3) The mathematical equation which is established by 16 groups of simulation data can characterize the causal relationship between the network traffic of specific phenomena and the spatial patterns of the group phenomena quantitatively.Through the above work, this article intends to reveal the spatial patterns of the public groups under networked geographic spaces from the experimental point of view, and provides a theoretical basis and methods in support of building and evolving of the spatial patterns of group phenomena under networked geographic spaces.