Community Detection and Analysis in Air Pollution Complex Network Based on LPA

Air pollution has caused a bad effect on the natural environment and public health. With the rapid establishment of air monitoring stations and meteorological stations, the volume of air pollution data is increasing dramatically. The traditional methods to model the air pollution transmission process have shown a limitation to process a large amount of non-linear data. Machine learning methods have recently been proven to be extremely efficient and accurate to process data in complex structures. In this paper, a novel method based on a complex network is proposed to discover the air monitoring station communities and analyze the air pollutants diffusion. The method comprises several aspects, including a dimension reduction process, a complex network generation process, and a community detection process. The proposed method was verified based on a real-world air pollution dataset, and the results demonstrate that the air pollutants diffusion has a community structure, and some critical air monitoring stations have a greater influence on the transmission of pollutants, which should be paid more attention to the following air quality management.