MANet: A Mining and Analysis Method of Air Pollutants Transmission Path Network

Air pollution is a serious environmental problem all over the world. Air has no specific form and pollutants transmission is affected by multi-dimensional factors. Hence, the path is difficult to describe. Based on the dynamic principle and causal mechanism, we propose MANet, an air pollutants transmission path network mining and analysis method integrating spatiotemporal factors and causal mechanism in historical data. Firstly, the pollutants monitoring stations are selected by grid method, while the valid data is screened through connection and balance. Secondly, the single source diffusion influence factor is defined. Key spatiotemporal influence factors are characterized and calculated, while the uncertain transmission path is measured from dynamic diffusion process. Thirdly, causal mechanism in historical data is mined and reasonable paths are screened. In the experiment, the monitoring data of pollutants concentration in Jing-Jin-Ji region is used to deeply explore the network performance forms, which proves the rationality of MANet and mine out the hidden rules of the network structure to provide guidance for air pollutants governance.