Hierarchical Clustering Algorithm In Research The Application Of The Network Society Division

In recent years, complex network is a hot international research topic, whosetheories widely applied to many fields of society. Constructing a completed datamodel has important theoretical significance on researching complex science,obtaining the features and functions of the network together with making full use ofthe network.With the wide range of applications of the complex network in various fields,network cooperated for scientific research, an example of the complex network, hasdeveloped into a network system of a certain scale. Both the evolutionary model andthe scientific collaboration network are important research contents.In this paper, the author built a small world network model based on the scientificdata collected from the researchers in a university and then analyzed it. The papermainly concentrates on applying the improved hierarchical clustering algorithm to thecommunity division of network in college scientific research. Each researcher in auniversity is considered as a node in the network. Thus we get a network model byconnecting two nodes which have partnerships with an edge. The model is analyzedby its clustering coefficient, the average path length and other characteristics vectors.The innovation of the paper is building a minimum spanning tree according to theedges’ degree coefficient. Firstly, a community hierarchical structure is built.Secondly, the community structure is portioned. Thirdly, the modularity maximizationis discovered. Finally, based on the analysis of modularity, the optimal community ofthe scientific collaboration network structure can be achieved.