Community Detection in Aviation Network Based on K-means and Complex Network

With the increasing number of airports and expansion their scale, aviation network has become complex hierarchical. In order to investigate characteristics networks, this paper constructs a Chinese model carries out related research based on theory K-means algorithm. Initially, P-space is employed construct model. Then, indicators such as degree, clustering coefficient, average path length, betweenness coreness are selected hierarchical features networks explore causes. Secondly, using algorithm, five values obtained initial parameter K for each hierarchies classified according indicators. Meanwhile, simulation experiments conducted obtain visual results nodes under different values, well silhouette coefficients evaluating effect indicator in classification Finally, coefficient optimal when value 4. Thus, four layers can be obtained. According experimental results, association discovery method combined with algorithm better applicability simplicity, while accuracy improved.